Recently, deep learning models have made great progress in MWP solving on answer accuracy. However, they are uninterpretable since they mainly rely on shallow heuristics to achieve high performance without understanding and reasoning the grounded math logic. To address this issue and make a step towards interpretable MWP solving, we first construct a high-quality MWP dataset named InterMWP which consists of 11,495 MWPs and annotates interpretable logical formulas based on algebraic knowledge as the grounded linguistic logic of each solution equation. Different from existing MWP datasets, our InterMWP benchmark asks for a solver to not only output the solution expressions but also predict the corresponding logical formulas. We further propose a novel approach with logical prompt and interpretation generation, called LogicSolver. For each MWP, our LogicSolver first retrieves some highly-correlated algebraic knowledge and then passes them to the backbone model as prompts to improve the semantic representations of MWPs. With these improved semantic representations, our LogicSolver generates corresponding solution expressions and interpretable knowledge formulas in accord with the generated solution expressions, simultaneously. Experimental results show that our LogicSolver has stronger logical formula-based interpretability than baselines while achieving higher answer accuracy with the help of logical prompts, simultaneously. The source code and dataset will be available at https://github.com/yangzhch6/InterMWP.
In e-commerce, the salience of commonsense knowledge (CSK) is beneficial for widespread applications such as product search and recommendation. For example, when users search for “running” in e-commerce, they would like to find products highly related to running, such as “running shoes” rather than “shoes”. Nevertheless, many existing CSK collections rank statements solely by confidence scores, and there is no information about which ones are salient from a human perspective. In this work, we define the task of supervised salience evaluation, where given a CSK triple, the model is required to learn whether the triple is salient or not. In addition to formulating the new task, we also release a new Benchmark dataset of Salience Evaluation in E-commerce (BSEE) and hope to promote related research on commonsense knowledge salience evaluation. We conduct experiments in the dataset with several representative baseline models. The experimental results show that salience evaluation is a hard task where models perform poorly on our evaluation set. We further propose a simple but effective approach, PMI-tuning, which shows promise for solving this novel problem. Code is available in https://github.com/OpenBGBenchmark/OpenBG-CSK.
Semi-supervised learning has shown promise in allowing NLP models to generalize from small amounts of labeled data. Meanwhile, pretrained transformer models act as black-box correlation engines that are difficult to explain and sometimes behave unreliably. In this paper, we propose tackling both of these challenges via Automatic Rule Induction (ARI), a simple and general-purpose framework for the automatic discovery and integration of symbolic rules into pretrained transformer models. First, we extract weak symbolic rules from low-capacity machine learning models trained on small amounts of labeled data. Next, we use an attention mechanism to integrate these rules into high-capacity pretrained transformer models. Last, the rule-augmented system becomes part of a self-training framework to boost supervision signal on unlabeled data. These steps can be layered beneath a variety of existing weak supervision and semi-supervised NLP algorithms in order to improve performance and interpretability. Experiments across nine sequence classification and relation extraction tasks suggest that ARI can improve state-of-the-art methods with no manual effort and minimal computational overhead.
Transformer-based pre-trained models like BERT have achieved great progress on Semantic Sentence Matching. Meanwhile, dependency prior knowledge has also shown general benefits in multiple NLP tasks. However, how to efficiently integrate dependency prior structure into pre-trained models to better model complex semantic matching relations is still unsettled. In this paper, we propose the Dependency-Enhanced Adaptive Fusion Attention (DAFA), which explicitly introduces dependency structure into pre-trained models and adaptively fuses it with semantic information. Specifically, (i) DAFA first proposes a structure-sensitive paradigm to construct a dependency matrix for calibrating attention weights. (ii) It adopts an adaptive fusion module to integrate the obtained dependency information and the original semantic signals. Moreover, DAFA reconstructs the attention calculation flow and provides better interpretability. By applying it on BERT, our method achieves state-of-the-art or competitive performance on 10 public datasets, demonstrating the benefits of adaptively fusing dependency structure in semantic matching task.
We combine the capacity of sparsely gated Mixture-of-Experts (MoE) with the speed and stability of linear, mixing transformations to design the Sparse Mixer encoder model. Sparse Mixer slightly outperforms BERT on GLUE and SuperGLUE, but more importantly trains 65% faster and runs inference 61% faster. We also present a faster variant, prosaically named Fast Sparse Mixer, that marginally underperforms BERT on SuperGLUE, but trains and runs nearly twice as fast. We justify the design of these two models by carefully ablating through various mixing mechanisms, MoE configurations, and hyperparameters. Sparse Mixer overcomes many of the latency and stability concerns of MoE models and offers the prospect of serving sparse student models, without resorting to distilling them to dense variants.
Commonsense question answering (CQA) aims to choose the correct answers for commonsense questions. Most existing works focus on extracting and reasoning over external knowledge graphs (KG). However, the noise in KG prevents these models from learning effective representations. In this paper, we propose a Knowledge Enhanced Graph Contrastive Learning model (KE-GCL) by incorporating the contextual descriptions of entities and adopting a graph contrastive learning scheme. Specifically, for QA pairs we represent the knowledge from KG and contextual descriptions. Then, the representations of contextual descriptions as context nodes are inserted into KG, forming the knowledge-enhanced graphs.Moreover, we design a contrastive learning method on graphs. For knowledge-enhanced graphs, we build their augmented views with an adaptive sampling strategy. After that, we reason over graphs to update their representations by scattering edges and aggregating nodes. To further improve GCL, hard graph negatives are chosen based on incorrect answers. Extensive experiments on two benchmark datasets demonstrate the effectiveness of our proposed KE-GCL, which outperforms previous methods consistently.
The role of the attention mechanism in encoding linguistic knowledge has received special interest in NLP. However, the ability of the attention heads to judge the grammatical acceptability of a sentence has been underexplored. This paper approaches the paradigm of acceptability judgments with topological data analysis (TDA), showing that the geometric properties of the attention graph can be efficiently exploited for two standard practices in linguistics: binary judgments and linguistic minimal pairs. Topological features enhance the BERT-based acceptability classifier scores by 8%-24% on CoLA in three languages (English, Italian, and Swedish). By revealing the topological discrepancy between attention maps of minimal pairs, we achieve the human-level performance on the BLiMP benchmark, outperforming nine statistical and Transformer LM baselines. At the same time, TDA provides the foundation for analyzing the linguistic functions of attention heads and interpreting the correspondence between the graph features and grammatical phenomena. We publicly release the code and other materials used in the experiments.
Derivative-free prompt learning has emerged as a lightweight alternative to prompt tuning, which only requires model inference to optimize the prompts. However, existing work did not take full advantage of the over-parameterized characteristics of large pre-trained language models (PLMs). In this paper, we propose Clip-Tuning, a simple yet effective method that adopts diverse frozen “thinned” networks of PLMs to obtain *a mixture of rewards* and thus advance the derivative-free prompt learning. The thinned networks consist of all the hidden units that survive a stationary dropout strategy, whose inference predictions reflect an ensemble of partial views over prompted training samples. Our method outperforms previous gradient-free prompt learning methods and achieves parity with gradient-based counterparts on seven language understanding benchmarks under few-shot settings.
Code contrastive pre-training has recently achieved significant progress on code-related tasks. In this paper, we present SCodeR, a Soft-labeled contrastive pre-training framework with two positive sample construction methods to learn functional-level CodeRepresentation. Considering the relevance between codes in a large-scale code corpus, the soft-labeled contrastive pre-training can obtain fine-grained soft-labels through an iterative adversarial manner and use them to learn better code representation. The positive sample construction is another key for contrastive pre-training. Previous works use transformation-based methods like variable renaming to generate semantically equal positive codes. However, they usually result in the generated code with a highly similar surface form, and thus mislead the model to focus on superficial code structure instead of code semantics. To encourage SCodeR to capture semantic information from the code, we utilize code comments and abstract syntax sub-trees of the code to build positive samples. We conduct experiments on four code-related tasks over seven datasets. Extensive experimental results show that SCodeR achieves new state-of-the-art performance on all of them, which illustrates the effectiveness of the proposed pre-training method.
Image-text retrieval is a fundamental cross-modal task that takes image/text as a query to retrieve relevant data of another type. The large-scale two-stream pre-trained models like CLIP have achieved tremendous success in this area. They embed the images and texts into instance representations with two separate encoders, aligning them on the instance-level with contrastive learning. Beyond this, the following works adopt the fine-grained token-level interaction (Masked Language and Image Modeling) to boost performance further. However, the vanilla token-level objectives are not designed to aggregate the image-text alignment information into the instance representations, but the token representations, causing a gap between pre-training and application. To address this issue, we carefully design two novel conditioned token-level pre-training objectives, Conditioned Masked Language and Image Modeling (ConMLM and ConMIM), forcing models to aggregate the token-level alignment information into the instance representations. Combing with the instance-level contrastive learning, we propose our cross-modal dense retrieval framework, Conditioned Language-Image Pre-training (ConLIP). Experimental results on two popular cross-modal retrieval benchmarks (MSCOCO and Flickr30k) reveal the effectiveness of our methods.
In simultaneous speech translation (SimulST), finding the best trade-off between high output quality and low latency is a challenging task. To meet the latency constraints posed by different application scenarios, multiple dedicated SimulST models are usually trained and maintained, generating high computational costs. In this paper, also motivated by the increased sensitivity towards sustainable AI, we investigate whether a single model trained offline can serve both offline and simultaneous applications under different latency regimes without additional training or adaptation. Experiments on en->de, es show that, aside from facilitating the adoption of well-established offline architectures and training strategies without affecting latency, offline training achieves similar or better quality compared to the standard SimulST training protocol, also being competitive with the state-of-the-art system.
Word translation without parallel corpora has become feasible, rivaling the performance of supervised methods. Recent findings have shown the improvement in accuracy and robustness of unsupervised word translation (UWT) by utilizing visual observations, which are universal representations across languages.Our work investigates the potential of using not only visual observations but also pretrained language-image models for enabling a more efficient and robust UWT. We develop a novel UWT method dubbed Word Alignment using Language-Image Pretraining (WALIP), leveraging visual observations via the shared image-text embedding space of CLIPs (Radford et al., 2021). WALIP has a two-step procedure. First, we retrieve word pairs with high confidences of similarity, computed using our proposed image-based fingerprints, which define the initial pivot for the alignment.Second, we apply our robust Procrustes algorithm to estimate the linear mapping between two embedding spaces, which iteratively corrects and refines the estimated alignment.Our extensive experiments show that WALIP improves upon the state-of-the-art performance of bilingual word alignment for a few language pairs across different word embeddings and displays great robustness to the dissimilarity of language pairs or training corpora for two word embeddings.
A common thread of open-domain question answering (QA) models employs a retriever-reader pipeline that first retrieves a handful of relevant passages from Wikipedia and then peruses the passages to produce an answer. However, even state-of-the-art readers fail to capture the complex relationships between entities appearing in questions and retrieved passages, leading to answers that contradict the facts. In light of this, we propose a novel knowledge graph enhanced passage reader, namely Grape, to improve the reader performance for open-domain QA. Specifically, for each pair of question and retrieved passage, we first construct a localized bipartite graph, attributed to entity embeddings extracted from the intermediate layer of the reader model. Then, a graph neural network learns relational knowledge while fusing graph and contextual representations into the hidden states of the reader model. Experiments on three open-domain QA benchmarks show Grape can improve the state-of-the-art performance by up to 2.2 exact match score with a negligible overhead increase, with the same retriever and retrieved passages. Our code is publicly available at https://github.com/jumxglhf/GRAPE.
Narrative summarization aims to produce a distilled version of a narrative to describe its most salient events and characters. Writing a summary for a narrative is challenging as it requires an understanding of event causality and character behaviors. To encourage research in this direction, we propose NarraSum, a large-scale narrative summarization dataset. It contains 122K narratives, which are collected from the synopses of movies and TV episodes with diverse genres, and their corresponding abstractive summaries. Experiments show that there is a large performance gap between humans and the state-of-the-art summarization models on NarraSum. We hope that this dataset will promote future research in summarization, as well as broader studies of natural language understanding and generation. The dataset is available at https://github.com/zhaochaocs/narrasum.
Being able to rank the similarity of short text segments is an interesting bonus feature of neural machine translation. Translation-based similarity measures include direct and pivot translation probability, as well as translation cross-likelihood, which has not been studied so far. We analyze these measures in the common framework of multilingual NMT, releasing the NMTScore library. Compared to baselines such as sentence embeddings, translation-based measures prove competitive in paraphrase identification and are more robust against adversarial or multilingual input, especially if proper normalization is applied. When used for reference-based evaluation of data-to-text generation in 2 tasks and 17 languages, translation-based measures show a relatively high correlation to human judgments.
Some claim language models understand us. Others won’t hear it. To clarify, I investigate three views of human language understanding: as-mapping, as-reliability and as-representation. I argue that while behavioral reliability is necessary for understanding, internal representations are sufficient; they climb the right hill. I review state-of-the-art language and multi-modal models: they are pragmatically challenged by under-specification of form. I question the Scaling Paradigm: limits on resources may prohibit scaled-up models from approaching understanding. Last, I describe how as-representation advances a science of understanding. We need work which probes model internals, adds more of human language, and measures what models can learn.
Dialogue meaning representation formulates natural language utterance semantics in their conversational context in an explicit and machine-readable form. Previous work typically follows the intent-slot framework, which is easy for annotation yet limited in scalability for complex linguistic expressions. A line of works alleviates the representation issue by introducing hierarchical structures but challenging to express complex compositional semantics, such as negation and coreference. We propose Dialogue Meaning Representation (DMR), a pliable and easily extendable representation for task-oriented dialogue. Our representation contains a set of nodes and edges to represent rich compositional semantics. Moreover, we propose an inheritance hierarchy mechanism focusing on domain extensibility. Additionally, we annotated DMR-FastFood, a multi-turn dialogue dataset with more than 70k utterances, with DMR. We propose two evaluation tasks to evaluate different dialogue models and a novel coreference resolution model GNNCoref for the graph-based coreference resolution task. Experiments show that DMR can be parsed well with pre-trained Seq2Seq models, and GNNCoref outperforms the baseline models by a large margin.The dataset and code are available at https://github.com/amazon-research/dialogue-meaning-representation
Chinese Spell Checking (CSC) aims to detect and correct Chinese spelling errors. Recent researches start from the pretrained knowledge of language models and take multimodal information into CSC models to improve the performance. However, they overlook the rich knowledge in the dictionary, the reference book where one can learn how one character should be pronounced, written, and used. In this paper, we propose the LEAD framework, which renders the CSC model to learn heterogeneous knowledge from the dictionary in terms of phonetics, vision, and meaning. LEAD first constructs positive and negative samples according to the knowledge of character phonetics, glyphs, and definitions in the dictionary. Then a unified contrastive learning-based training scheme is employed to refine the representations of the CSC models. Extensive experiments and detailed analyses on the SIGHAN benchmark datasets demonstrate the effectiveness of our proposed methods.
Despite their recent popularity and well-known advantages, dense retrievers still lag behind sparse methods such as BM25 in their ability to reliably match salient phrases and rare entities in the query and to generalize to out-of-domain data. It has been argued that this is an inherent limitation of dense models. We rebut this claim by introducing the Salient Phrase Aware Retriever (SPAR), a dense retriever with the lexical matching capacity of a sparse model. We show that a dense Lexical Model Λ can be trained to imitate a sparse one, and SPAR is built by augmenting a standard dense retriever with Λ. Empirically, SPAR shows superior performance on a range of tasks including five question answering datasets, MS MARCO passage retrieval, as well as the EntityQuestions and BEIR benchmarks for out-of-domain evaluation, exceeding the performance of state-of-the-art dense and sparse retrievers. The code and models of SPAR are available at: https://github.com/facebookresearch/dpr-scale/tree/main/spar
Automatic product attribute value extraction refers to the task of identifying values of an attribute from the product information. Product attributes are essential in improving online shopping experience for customers. Most existing methods focus on extracting attribute values from product title and description.However, in many real-world applications, a product is usually represented by multiple modalities beyond title and description, such as product specifications, text and visual information from the product image, etc. In this paper, we propose SMARTAVE, a Structure Mltimodal trAnsformeR for producT Attribute Value Extraction, which jointly encodes the structured product information from multiple modalities. Specifically, in SMARTAVE encoder, we introduce hyper-tokens to represent the modality-level information, and local-tokens to represent the original text and visual inputs. Structured attention patterns are designed among the hyper-tokens and local-tokens for learning effective product representation. The attribute values are then extracted based on the learned embeddings. We conduct extensive experiments on two multimodal product datasets. Experimental results demonstrate the superior performance of the proposed approach over several state-of-the-art methods. Ablation studies validate the effectiveness of the structured attentions in modeling the multimodal product information.
With the rapid development of pre-training techniques, a number of language models have been pre-trained on large-scale code corpora and perform well in code generation. In this paper, we investigate how to equip pre-trained language models with the ability of code generation for private libraries. In practice, it is common for programmers to write code using private libraries. However, this is a challenge for language models since they have never seen private APIs during training. Motivated by the fact that private libraries usually come with elaborate API documentation, we propose a novel framework with two modules: the APIRetriever finds useful APIs, and then the APICoder generates code using these APIs. For APIRetriever, we present a dense retrieval system and also design a friendly interaction to involve uses. For APICoder, we can directly use off-the-shelf language models, or continually pre-train the base model on a code corpus containing API information. Both modules are trained with data from public libraries and can be generalized to private ones. Furthermore, we craft three benchmarks for private libraries, named TorchDataEval, MonkeyEval, and BeatNumEval. Experimental results demonstrate the impressive performance of our framework.
Previous studies on cross-domain sentiment classification depend on the pivot features or utilize the target data for representation learning, which ignore the semantic relevance between different domains. To this end, we exploit Abstract Meaning Representation (AMR) to help with cross-domain sentiment classification. Compared with the textual input, AMR reduces data sparsity and explicitly provides core semantic knowledge and correlations between different domains. In particular, we develop an algorithm to construct a sentiment-driven semantic graph from sentence-level AMRs. We further design two strategies to linearize the semantic graph and propose a text-graph interaction model to fuse the text and semantic graph representations for cross-domain sentiment classification. Empirical studies show the effectiveness of our proposed model over several strong baselines. The results also indicate the importance of the proposed sentiment-driven semantic graph for cross-domain sentiment classification.
The shift towards publicly available text sources has enabled language processing at unprecedented scale, yet leaves under-serviced the domains where public and openly licensed data is scarce. Proactively collecting text data for research is a viable strategy to address this scarcity, but lacks systematic methodology taking into account the many ethical, legal and confidentiality-related aspects of data collection. Our work presents a case study on proactive data collection in peer review – a challenging and under-resourced NLP domain. We outline ethical and legal desiderata for proactive data collection and introduce “Yes-Yes-Yes”, the first donation-based peer reviewing data collection workflow that meets these requirements. We report on the implementation of Yes-Yes-Yes at ACL Rolling Review and empirically study the implications of proactive data collection for the dataset size and the biases induced by the donation behavior on the peer reviewing platform.
It is still a pipe dream that personal AI assistants on the phone and AR glasses can assist our daily life in addressing our questions like “how to adjust the date for this watch?” and “how to set its heating duration? (while pointing at an oven)”. The queries used in conventional tasks (i.e. Video Question Answering, Video Retrieval, Moment Localization) are often factoid and based on pure text. In contrast, we present a new task called Task-oriented Question-driven Video Segment Retrieval (TQVSR). Each of our questions is an image-box-text query that focuses on affordance of items in our daily life and expects relevant answer segments to be retrieved from a corpus of instructional video-transcript segments. To support the study of this TQVSR task, we construct a new dataset called AssistSR. We design novel guidelines to create high-quality samples. This dataset contains 3.2k multimodal questions on 1.6k video segments from instructional videos on diverse daily-used items. To address TQVSR, we develop a simple yet effective model called Dual Multimodal Encoders (DME) that significantly outperforms several baseline methods while still having large room for improvement in the future. Moreover, we present detailed ablation analyses. Code and data are available at https://github.com/StanLei52/TQVSR.
Despite the potential of federated learning, it is known to be vulnerable to backdoor attacks. Many robust federated aggregation methods are proposed to reduce the potential backdoor risk. However, they are mainly validated in the CV field. In this paper, we find that NLP backdoors are hard to defend against than CV, and we provide a theoretical analysis that the malicious update detection error probabilities are determined by the relative backdoor strengths. NLP attacks tend to have small relative backdoor strengths, which may result in the failure of robust federated aggregation methods for NLP attacks. Inspired by the theoretical results, we can choose some dimensions with higher backdoor strengths to settle this issue. We propose a novel federated aggregation algorithm, Dim-Krum, for NLP tasks, and experimental results validate its effectiveness.
Deep Neural Networks (DNNs) are known to be vulnerable to backdoor attacks. In Natural Language Processing (NLP), DNNs are often backdoored during the fine-tuning process of a large-scale Pre-trained Language Model (PLM) with poisoned samples. Although the clean weights of PLMs are readily available, existing methods have ignored this information in defending NLP models against backdoor attacks. In this work, we take the first step to exploit the pre-trained (unfine-tuned) weights to mitigate backdoors in fine-tuned language models. Specifically, we leverage the clean pre-trained weights via two complementary techniques: (1) a two-step Fine-mixing technique, which first mixes the backdoored weights (fine-tuned on poisoned data) with the pre-trained weights, then fine-tunes the mixed weights on a small subset of clean data; (2) an Embedding Purification (E-PUR) technique, which mitigates potential backdoors existing in the word embeddings. We compare Fine-mixing with typical backdoor mitigation methods on three single-sentence sentiment classification tasks and two sentence-pair classification tasks and show that it outperforms the baselines by a considerable margin in all scenarios. We also show that our E-PUR method can benefit existing mitigation methods. Our work establishes a simple but strong baseline defense for secure fine-tuned NLP models against backdoor attacks.
Language models (LMs) have recently been shown to generate more factual responses by employing modularity (Zhou et al., 2022) in combination with retrieval (Adolphs et al., 2021). We extend the recent approach of Adolphs et al. (2021) to include internet search as a module. Our SeeKeR (Search engine->Knowledge->Response) method thus applies a single LM to three modular tasks in succession: search, generating knowledge, and generating a final response. We show that, when using SeeKeR as a dialogue model, it outperforms the state-of-the-art model BlenderBot 2 (Chen et al., 2021) on open-domain knowledge-grounded conversations for the same number of parameters, in terms of consistency, knowledge and per-turn engagingness. SeeKeR applied to topical prompt completions as a standard language model outperforms GPT2 (Radford et al., 2019) and GPT3 (Brown et al., 2020) in terms of factuality and topicality, despite GPT3 being a vastly larger model. Our code and models are made publicly available.
Natural Language Inference (NLI) has been extensively studied by the NLP community as a framework for estimating the semantic relation between sentence pairs. While early work identified certain biases in NLI models, recent advancements in modeling and datasets demonstrated promising performance.In this work, we further explore the direct zero-shot applicability of NLI models to real applications, beyond the sentence-pair setting they were trained on. First, we analyze the robustness of these models to longer and out-of-domain inputs. Then, we develop new aggregation methods to allow operating over full documents, reaching state-of-the-art performance on the ContractNLI dataset. Interestingly, we find NLI scores to provide strong retrieval signals, leading to more relevant evidence extractions compared to common similarity-based methods. Finally, we go further and investigate whole document clusters to identify both discrepancies and consensus among sources. In a test case, we find real inconsistencies between Wikipedia pages in different languages about the same topic.
This paper presents an empirical study to build relation extraction systems in low-resource settings. Based upon recent pre-trained language models, we comprehensively investigate three schemes to evaluate the performance in low-resource settings: (i) different types of prompt-based methods with few-shot labeled data; (ii) diverse balancing methods to address the long-tailed distribution issue; (iii) data augmentation technologies and self-training to generate more labeled in-domain data. We create a benchmark with 8 relation extraction (RE) datasets covering different languages, domains and contexts and perform extensive comparisons over the proposed schemes with combinations. Our experiments illustrate: (i) Though prompt-based tuning is beneficial in low-resource RE, there is still much potential for improvement, especially in extracting relations from cross-sentence contexts with multiple relational triples; (ii) Balancing methods are not always helpful for RE with long-tailed distribution; (iii) Data augmentation complements existing baselines and can bring much performance gain, while self-training may not consistently achieve advancement to low-resource RE. Code and datasets are in https://github.com/zjunlp/LREBench.
Continual Language Learning (CLL) in multilingual translation is inevitable when new languages are required to be translated. Due to the lack of unified and generalized benchmarks, the evaluation of existing methods is greatly influenced by experimental design which usually has a big gap from the industrial demands. In this work, we propose the first Continual Language Learning Evaluation benchmark CLLE in multilingual translation. CLLE consists of a Chinese-centric corpus — CN-25 and two CLL tasks — the close-distance language continual learning task and the language family continual learning task designed for real and disparate demands. Different from existing translation benchmarks, CLLE considers several restrictions for CLL, including domain distribution alignment, content overlap, language diversity, and the balance of corpus. Furthermore, we propose a novel framework COMETA based on Constrained Optimization and META-learning to alleviate catastrophic forgetting and dependency on history training data by using a meta-model to retain the important parameters for old languages. Our experiments prove that CLLE is a challenging CLL benchmark and that our proposed method is effective when compared with other strong baselines. Due to the construction of the corpus, the task designing and the evaluation method are independent of the centric language, we also construct and release the English-centric corpus EN-25 to facilitate academic research.
Recent multilingual pre-trained models have shown better performance in various multilingual tasks. However, these models perform poorly on multilingual retrieval tasks due to lacking multilingual training data. In this paper, we propose to mine and generate self-supervised training data based on a large-scale unlabeled corpus. We carefully design a mining method which combines the sparse and dense models to mine the relevance of unlabeled queries and passages. And we introduce a query generator to generate more queries in target languages for unlabeled passages. Through extensive experiments on Mr. TYDI dataset and an industrial dataset from a commercial search engine, we demonstrate that our method performs better than baselines based on various pre-trained multilingual models. Our method even achieves on-par performance with the supervised method on the latter dataset.
Although explainable artificial intelligence (XAI) has achieved remarkable developments in recent years, there are few efforts have been devoted to the following problems, namely, i) how to develop an explainable method that could explain the black-box in a model-agnostic way? and ii) how to improve the performance and interpretability of the black-box using such explanations instead of pre-collected important attributions? To explore the potential solution, we propose a model-agnostic explanation method termed as Sparse Contrastive Coding (SCC) and verify its effectiveness in text classification and natural language inference. In brief, SCC explains the feature attributions which characterize the importance of words based on the hidden states of each layer of the model. With such word-level explainability, SCC adaptively divides the input sentences into foregrounds and backgrounds in terms of task relevance. Through maximizing the similarity between the foregrounds and input sentences while minimizing the similarity between the backgrounds and input sentences, SSC employs a supervised contrastive learning loss to boost the interpretability and performance of the model. Extensive experiments show the superiority of our method over five state-of-the-art methods in terms of interpretability and classification measurements. The code is available at https://pengxi.me.
Language-based environment manipulation requires agents to manipulate the environment following natural language instructions, which is challenging due to the huge space of the environments.To address this challenge, various approaches have been proposed in recent work. Although these approaches work well for their intended environments, they are difficult to generalize across environments. In this work, we propose LEMON, a general framework for language-based environment manipulation tasks. Specifically, we first specify a general approach for language-based environment manipulation tasks, which can deal with various environments using the same generative language model. Then we propose an execution-guided pre-training strategy to inject prior knowledge of environments to the language model with a pure synthetic pre-training corpus. Experimental results on tasks including Alchemy, Scene, Tangrams, ProPara and Recipes demonstrate the effectiveness of LEMON: it achieves new state-of-the-art results on four of the tasks, and the execution-guided pre-training strategy brings remarkable improvements on all experimental tasks.
Named entity recognition (NER) suffers from the scarcity of annotated training data, especially for low-resource languages without labeled data. Cross-lingual NER has been proposed to alleviate this issue by transferring knowledge from high-resource languages to low-resource languages via aligned cross-lingual representations or machine translation results. However, the performance of cross-lingual NER methods is severely affected by the unsatisfactory quality of translation or label projection. To address these problems, we propose a Cross-lingual Entity Projection framework (CROP) to enable zero-shot cross-lingual NER with the help of a multilingual labeled sequence translation model. Specifically, the target sequence is first translated into the source language and then tagged by a source NER model. We further adopt a labeled sequence translation model to project the tagged sequence back to the target language and label the target raw sentence. Ultimately, the whole pipeline is integrated into an end-to-end model by the way of self-training. Experimental results on two benchmarks demonstrate that our method substantially outperforms the previous strong baseline by a large margin of +3 7 F1 scores and achieves state-of-the-art performance.
Text data can pose a risk of harm. However, the risks are not fully understood, and how to handle, present, and discuss harmful text in a safe way remains an unresolved issue in the NLP community. We provide an analytical framework categorising harms on three axes: (1) the harm type (e.g., misinformation, hate speech or racial stereotypes); (2) whether a harm is sought as a feature of the research design if explicitly studying harmful content (e.g., training a hate speech classifier), versus unsought if harmful content is encountered when working on unrelated problems (e.g., language generation or part-of-speech tagging); and (3) who it affects, from people (mis)represented in the data to those handling the data and those publishing on the data. We provide advice for practitioners, with concrete steps for mitigating harm in research and in publication. To assist implementation we introduce HarmCheck – a documentation standard for handling and presenting harmful text in research.
Multimodal representation learning is a challenging task in which previous work mostly focus on either uni-modality pre-training or cross-modality fusion. In fact, we regard modeling multimodal representation as building a skyscraper, where laying stable foundation and designing the main structure are equally essential. The former is like encoding robust uni-modal representation while the later is like integrating interactive information among different modalities, both of which are critical to learning an effective multimodal representation. Recently, contrastive learning has been successfully applied in representation learning, which can be utilized as the pillar of the skyscraper and benefit the model to extract the most important features contained in the multimodal data. In this paper, we propose a novel framework named MultiModal Contrastive Learning (MMCL) for multimodal representation to capture intra- and inter-modality dynamics simultaneously. Specifically, we devise uni-modal contrastive coding with an efficient uni-modal feature augmentation strategy to filter inherent noise contained in acoustic and visual modality and acquire more robust uni-modality representations. Besides, a pseudo siamese network is presented to predict representation across different modalities, which successfully captures cross-modal dynamics. Moreover, we design two contrastive learning tasks, instance- and sentiment-based contrastive learning, to promote the process of prediction and learn more interactive information related to sentiment. Extensive experiments conducted on two public datasets demonstrate that our method surpasses the state-of-the-art methods.
Prompt-based fine-tuning has boosted the performance of Pre-trained Language Models (PLMs) on few-shot text classification by employing task-specific prompts. Yet, PLMs are unfamiliar with prompt-style expressions during pre-training, which limits the few-shot learning performance on downstream tasks.It would be desirable if the models can acquire some prompting knowledge before adapting to specific NLP tasks. We present the Unified Prompt Tuning (UPT) framework, leading to better few-shot text classification for BERT-style models by explicitly capturing prompting semantics from non-target NLP datasets. In UPT, a novel paradigm Prompt-Options-Verbalizer is proposed for joint prompt learning across different NLP tasks, forcing PLMs to capture task-invariant prompting knowledge. We further design a self-supervised task named Knowledge-enhanced Selective Masked Language Modeling to improve the PLM’s generalization abilities for accurate adaptation to previously unseen tasks. After multi-task learning across multiple tasks, the PLM can be better prompt-tuned towards any dissimilar target tasks in low-resourced settings. Experiments over a variety of NLP tasks show that UPT consistently outperforms state-of-the-arts for prompt-based fine-tuning.
Language Models (LMs) can perform new tasks by adapting to a few in-context examples. For humans, explanations that connect examples to task principles can improve learning. We therefore investigate whether explanations of few-shot examples can help LMs. We annotate questions from 40 challenging tasks with answer explanations, and various matched control explanations. We evaluate how different types of explanations, instructions, and controls affect zero- and few-shot performance. We analyze these results using statistical multilevel modeling techniques that account for the nested dependencies among conditions, tasks, prompts, and models. We find that explanations can improve performance—even without tuning. Furthermore, explanations hand-tuned for performance on a small validation set offer substantially larger benefits, and building a prompt by selecting examples and explanations together substantially improves performance over selecting examples alone. Finally, even untuned explanations outperform carefully matched controls, suggesting that the benefits are due to the link between an example and its explanation, rather than lower-level features. However, only large models benefit. In summary, explanations can support the in-context learning of large LMs on challenging tasks.
Recently, retrieval models based on dense representations are dominant in passage retrieval tasks, due to their outstanding ability in terms of capturing semantics of input text compared to the traditional sparse vector space models. A common practice of dense retrieval models is to exploit a dual-encoder architecture to represent a query and a passage independently. Though efficient, such a structure loses interaction between the query-passage pair, resulting in inferior accuracy. To enhance the performance of dense retrieval models without loss of efficiency, we propose a GNN-encoder model in which query (passage) information is fused into passage (query) representations via graph neural networks that are constructed by queries and their top retrieved passages. By this means, we maintain a dual-encoder structure, and retain some interaction information between query-passage pairs in their representations, which enables us to achieve both efficiency and efficacy in passage retrieval. Evaluation results indicate that our method significantly outperforms the existing models on MSMARCO, Natural Questions and TriviaQA datasets, and achieves the new state-of-the-art on these datasets.
Chinese Grammatical Error Correction (CGEC) is both a challenging NLP task and a common application in human daily life. Recently, many data-driven approaches are proposed for the development of CGEC research. However, there are two major limitations in the CGEC field: First, the lack of high-quality annotated training corpora prevents the performance of existing CGEC models from being significantly improved. Second, the grammatical errors in widely used test sets are not made by native Chinese speakers, resulting in a significant gap between the CGEC models and the real application. In this paper, we propose a linguistic rules-based approach to construct large-scale CGEC training corpora with automatically generated grammatical errors. Additionally, we present a challenging CGEC benchmark derived entirely from errors made by native Chinese speakers in real-world scenarios. Extensive experiments and detailed analyses not only demonstrate that the training data constructed by our method effectively improves the performance of CGEC models, but also reflect that our benchmark is an excellent resource for further development of the CGEC field.
Temporal sentence grounding (TSG) aims to identify the temporal boundary of a specific segment from an untrimmed video by a sentence query. All existing works first utilize a sparse sampling strategy to extract a fixed number of video frames and then interact them with query for reasoning.However, we argue that these methods have overlooked two indispensable issues:1) Boundary-bias: The annotated target segment generally refers to two specific frames as corresponding start and end timestamps. The video downsampling process may lose these two frames and take the adjacent irrelevant frames as new boundaries.2) Reasoning-bias: Such incorrect new boundary frames also lead to the reasoning bias during frame-query interaction, reducing the generalization ability of model.To alleviate above limitations, in this paper, we propose a novel Siamese Sampling and Reasoning Network (SSRN) for TSG, which introduces a siamese sampling mechanism to generate additional contextual frames to enrich and refine the new boundaries. Specifically, a reasoning strategy is developed to learn the inter-relationship among these frames and generate soft labels on boundaries for more accurate frame-query reasoning. Such mechanism is also able to supplement the absent consecutive visual semantics to the sampled sparse frames for fine-grained activity understanding.Extensive experiments demonstrate the effectiveness of SSRN on three challenging datasets.
Logical reasoning is a challenge for many current NLP neural network models since it requires more than the ability of learning informative representations from data. Inspired by the Dual Process Theory in cognitive science — which proposes that human cognition process involves two stages: an intuitive, unconscious and fast process relying on perception calledSystem 1, and a logical, conscious and slow process performing complex reasoning called System 2 — we leverage neural logic reasoning (System 2) on top of the representation learning models (System 1), which conducts explicit neural-based differentiable logical reasoning on top of the representations learned by the base neural models. Based on experiments on the commonsense knowledge graph completion task, we show that the two-system architecture always improves from its System 1 model alone. Experiments also show that both the rule-driven logical regularizer and the data-driven value regularizer are important and the performance improvement is marginal without the two regularizers, which indicates that learning from both logical prior and training data is important for reasoning tasks.
Federated learning (FL) can be essential in knowledge representation, reasoning, and data mining applications over multi-source knowledge graphs (KGs). A recent study FedE first proposes an FL framework that shares entity embeddings of KGs across all clients. However, entity embedding sharing from FedE would incur a severe privacy leakage. Specifically, the known entity embedding can be used to infer whether a specific relation between two entities exists in a private client. In this paper, we introduce a novel attack method that aims to recover the original data based on the embedding information, which is further used to evaluate the vulnerabilities of FedE. Furthermore, we propose a Federated learning paradigm with privacy-preserving Relation embedding aggregation (FedR) to tackle the privacy issue in FedE. Besides, relation embedding sharing can significantly reduce the communication cost due to its smaller size of queries. We conduct extensive experiments to evaluate FedR with five different KG embedding models and three datasets. Compared to FedE, FedR achieves similar utility and significant improvements regarding privacy-preserving effect and communication efficiency on the link prediction task.
Existing textual adversarial attacks usually utilize the gradient or prediction confidence to generate adversarial examples, making it hard to be deployed in real-world applications. To this end, we consider a rarely investigated but more rigorous setting, namely hard-label attack, in which the attacker can only access the prediction label. In particular, we find we can learn the importance of different words via the change on prediction label caused by word substitutions on the adversarial examples. Based on this observation, we propose a novel adversarial attack, termed Text Hard-label attacker (TextHacker). TextHacker randomly perturbs lots of words to craft an adversarial example. Then, TextHacker adopts a hybrid local search algorithm with the estimation of word importance from the attack history to minimize the adversarial perturbation. Extensive evaluations for text classification and textual entailment show that TextHacker significantly outperforms existing hard-label attacks regarding the attack performance as well as adversary quality.
Neural language models encode rich knowledge about entities and their relationships which can be extracted from their representations using probing. Common properties of nouns (e.g., red strawberries, small ant) are, however, more challenging to extract compared to other types of knowledge because they are rarely explicitly stated in texts.We hypothesize this to mainly be the case for perceptual properties which are obvious to the participants in the communication. We propose to extract these properties from images and use them in an ensemble model, in order to complement the information that is extracted from language models. We consider perceptual properties to be more concrete than abstract properties (e.g., interesting, flawless). We propose to use the adjectives’ concreteness score as a lever to calibrate the contribution of each source (text vs. images). We evaluate our ensemble model in a ranking task where the actual properties of a noun need to be ranked higher than other non-relevant properties. Our results show that the proposed combination of text and images greatly improves noun property prediction compared to powerful text-based language models.
Metaphor Detection aims to identify the metaphorical meaning of words in the sentence. Most existing work is discriminant models, which use the contextual semantic information extracted by transformers for classifications directly. Due to insufficient training data and corresponding paraphrases, recent methods focus on how to get external resources and utilize them to introduce more knowledge. Currently, contextual modeling and external data are two key issues in the field. In this paper, we propose **A**n **A**uto-**A**ugmented **S**tructure-aware generative model (**AAAS**) for metaphor detection, which transforms the classification task into a keywords-extraction task. Specifically, we propose the task of structure information extraction to allow the model to use the ‘structural language’ to describe the whole sentence. Furthermore, without any other external resources, we design a simple but effective auto-augmented method to expand the limited datasets. Experimental results show that **AAAS** obtains competitive results compared with state-of-the-art methods.
Natural language processing (NLP) models are known to be vulnerable to backdoor attacks, which poses a newly arisen threat to NLP models. Prior online backdoor defense methods for NLP models only focus on the anomalies at either the input or output level, still suffering from fragility to adaptive attacks and high computational cost. In this work, we take the first step to investigate the unconcealment of textual poisoned samples at the intermediate-feature level and propose a feature-based efficient online defense method. Through extensive experiments on existing attacking methods, we find that the poisoned samples are far away from clean samples in the intermediate feature space of a poisoned NLP model. Motivated by this observation, we devise a distance-based anomaly score (DAN) to distinguish poisoned samples from clean samples at the feature level. Experiments on sentiment analysis and offense detection tasks demonstrate the superiority of DAN, as it substantially surpasses existing online defense methods in terms of defending performance and enjoys lower inference costs. Moreover, we show that DAN is also resistant to adaptive attacks based on feature-level regularization. Our code is available at https://github.com/lancopku/DAN.
Knowledge Graph(KG) grounded conversations often use large pre-trained models and usually suffer from fact hallucination. Frequently entities with no references in knowledge sources and conversation history are introduced into responses, thus hindering the flow of the conversation—existing work attempt to overcome this issue by tweaking the training procedure or using a multi-step refining method. However, minimal effort is put into constructing an entity-level hallucination detection system, which would provide fine-grained signals that control fallacious content while generating responses. As a first step to address this issue, we dive deep to identify various modes of hallucination in KG-grounded chatbots through human feedback analysis. Secondly, we propose a series of perturbation strategies to create a synthetic dataset named FADE (FActual Dialogue Hallucination DEtection Dataset). Finally, we conduct comprehensive data analyses and create multiple baseline models for hallucination detection to compare against human-verified data and already established benchmarks.
State-of-the-art keyphrase generation methods generally depend on large annotated datasets, limiting their performance in domains with limited annotated data. To overcome this challenge, we design a data-oriented approach that first identifies salient information using retrieval-based corpus-level statistics, and then learns a task-specific intermediate representation based on a pre-trained language model using large-scale unlabeled documents. We introduce salient span recovery and salient span prediction as denoising training objectives that condense the intra-article and inter-article knowledge essential for keyphrase generation. Through experiments on multiple keyphrase generation benchmarks, we show the effectiveness of the proposed approach for facilitating low-resource keyphrase generation and zero-shot domain adaptation. Our method especially benefits the generation of absent keyphrases, approaching the performance of models trained with large training sets.
Levin et al. (2019) show experimentally that the interpretations of novel English noun compounds (e.g., stew skillet), while not fully compositional, are highly predictable based on whether the modifier and head refer to artifacts or natural kinds. Is the large language model GPT-3 governed by the same interpretive principles? To address this question, we first compare Levin et al.’s experimental data with GPT-3 generations, finding a high degree of similarity. However, this evidence is consistent with GPT-3 reasoning only about specific lexical items rather than the more abstract conceptual categories of Levin et al.’s theory. To probe more deeply, we construct prompts that require the relevant kind of conceptual reasoning. Here, we fail to find convincing evidence that GPT-3 is reasoning about more than just individual lexical items. These results highlight the importance of controlling for low-level distributional regularities when assessing whether a large language model latently encodes a deeper theory.
Recent approaches to empathetic response generation incorporate emotion causalities to enhance comprehension of both the user’s feelings and experiences. However, these approaches suffer from two critical issues. First, they only consider causalities between the user’s emotion and the user’s experiences, and ignore those between the user’s experiences. Second, they neglect interdependence among causalities and reason them independently. To solve the above problems, we expect to reason all plausible causalities interdependently and simultaneously, given the user’s emotion, dialogue history, and future dialogue content. Then, we infuse these causalities into response generation for empathetic responses. Specifically, we design a new model, i.e., the Conditional Variational Graph Auto-Encoder (CVGAE), for the causality reasoning, and adopt a multi-source attention mechanism in the decoder for the causality infusion. We name the whole framework as CARE, abbreviated for CAusality Reasoning for Empathetic conversation. Experimental results indicate that our method achieves state-of-the-art performance.
Zero-Resource Cross-Lingual Named Entity Recognition aims at training an NER model of the target language using only labeled source language data and unlabeled target language data. Existing methods are mainly divided into three categories: model transfer based, data transfer based and knowledge transfer based. Each method has its own disadvantages, and combining more than one of them often leads to better performance. However, the performance of data transfer based methods is often limited by inevitable noise in the translation process. To handle the problem, we propose a framework named TransAdv to mitigate lexical and syntactic errors of word-by-word translated data, better utilizing the data by multi-level adversarial learning and multi-model knowledge distillation. Extensive experiments are conducted over 6 target languages with English as the source language, and the results show that TransAdv achieves competitive performance to the state-of-the-art models.
Machine learning models often suffer from a performance drop when they are applied to out-of-distribution (OOD) samples, i.e., those drawn far away from the training data distribution. Existing OOD detection work mostly focuses on identifying semantic-shift OOD samples, e.g., instances from unseen new classes. However, background-shift OOD detection, which identifies samples with domain or style-change, represents a more practical yet challenging task. In this paper, we propose Background-Aware Representation Learning (BARLE) for background-shift OOD detection in NLP. Specifically, we generate semantics-preserving background-shifted pseudo OOD samples from pretrained masked language models. We then contrast the in-distribution (ID) samples with their pseudo OOD counterparts. Unlike prior semantic-shift OOD detection work that often leverages an external text corpus, BARLE only uses ID data, which is more flexible and cost-efficient. In experiments across several text classification tasks, we demonstrate that BARLE is capable of improving background-shift OOD detection performance while maintaining ID classification accuracy. We further investigate the properties of the generated pseudo OOD samples, uncovering the working mechanism of BARLE.
The crystallization of modeling methods around the Transformer architecture has been a boon for practitioners. Simple, well-motivated architectural variations can transfer across tasks and scale, increasing the impact of modeling research. However, with the emergence of state-of-the-art 100B+ parameters models, large language models are increasingly expensive to accurately design and train. Notably, it can be difficult to evaluate how modeling decisions may impact emergent capabilities, given that these capabilities arise mainly from sheer scale alone.In the process of building BLOOM–the Big Science Large Open-science Open-access Multilingual language model–our goal is to identify an architecture and training setup that makes the best use of our 1,000,000 A100-GPU-hours budget.Specifically, we perform an ablation study at the billion-parameter scale comparing different modeling practices and their impact on zero-shot generalization.In addition, we study the impact of various popular pre-training corpora on zero-shot generalization. We also study the performance of a multilingual model and how it compares to the English-only one. Finally, we consider the scaling behaviour of Transformers to choose the target model size, shape, and training setup. All our models and code are open-sourced at https://huggingface.co/bigscience.
Out-of-distribution (OOD) detection aims to discern outliers from the intended data distribution, which is crucial to maintaining high reliability and a good user experience.Most recent studies in OOD detection utilize the information from a single representation that resides in the penultimate layer to determine whether the input is anomalous or not.Although such a method is straightforward, the potential of diverse information in the intermediate layers is overlooked.In this paper, we propose a novel framework based on contrastive learning that encourages intermediate features to learn layer-specialized representations and assembles them implicitly into a single representation to absorb rich information in the pre-trained language model. Extensive experiments in various intent classification and OOD datasets demonstrate that our approach is significantly more effective than other works.
Pretrained language models can be effectively stimulated by textual prompts or demonstrations, especially in low-data scenarios. Recent works have focused on automatically searching discrete or continuous prompts or optimized verbalizers, yet studies for the demonstration are still limited. Concretely, the demonstration examples are crucial for an excellent final performance of prompt-tuning. In this paper, we propose a novel pluggable, extensible, and efficient approach named contrastive demonstration tuning, which is free of demonstration sampling. Furthermore, the proposed approach can be: (i) Plugged into any previous prompt-tuning approaches; (ii) Extended to widespread classification tasks with a large number of categories. Experimental results on 16 datasets illustrate that our method integrated with previous approaches LM-BFF and P-tuning can yield better performance. Code is available in https://github.com/zjunlp/PromptKG/tree/main/research/Demo-Tuning.
Automated software debugging is a crucial task for improving the productivity of software developers. Many neural-based techniques have been proven effective for debugging-related tasks such as bug localization and program repair (or bug fixing). However, these techniques often focus only on either one of them or approach them in a stage-wise manner, ignoring the mutual benefits between them. In this work, we propose a novel unified Detect-Localize-Repair framework based on a pretrained programming language model CodeT5 to seamlessly address these tasks, named CodeT5-DLR. Specifically, we propose three objectives to adapt the generic CodeT5 for debugging: a bug detection objective to determine whether a given code snippet is buggy or not, a bug localization objective to identify the buggy lines, and a program repair objective to translate the buggy code to its fixed version. We evaluate it on each of these tasks and their combined setting on two newly collected line-level debugging datasets in Java and Python. Extensive results show that our model significantly outperforms existing baselines from both NLP and software engineering domains.
Many standard tasks in NLP (e.g., Named Entity Recognition, Part-of-Speech tagging, and Semantic Role Labeling) are naturally framed as sequence tagging problems. However, there has been comparatively little work on interpretability methods for sequence tagging models. In this paper, we extend influence functions — which aim to trace predictions back to the training points that informed them — to sequence tagging tasks. We define the influence of a training instance segment as the effect that perturbing the labels within this segment has on a test segment level prediction. We provide an efficient approximation to compute this, and show that it tracks with the “true” segment influence (measured empirically). We show the practical utility of segment influence by using the method to identify noisy annotations in NER corpora.
Pretrained Language Models (LMs) have demonstrated ability to perform numerical reasoning by extrapolating from a few examples in few-shot settings. However, the extent to which this extrapolation relies on robust reasoning is unclear. In this paper, we investigate how well these models reason with terms that are less frequent in the pretraining data. In particular, we examine the correlations between the model performance on test instances and the frequency of terms from those instances in the pretraining data. We measure the strength of this correlation for a number of GPT-based language models (pretrained on the Pile dataset) on various numerical deduction tasks (e.g., arithmetic and unit conversion). Our results consistently demonstrate that models are more accurate on instances whose terms are more prevalent, in some cases above 70% (absolute) more accurate on the top 10% frequent terms in comparison to the bottom 10%. Overall, although LMs appear successful at few-shot numerical reasoning, our results raise the question of how much models actually generalize beyond pretraining data, and we encourage researchers to take the pretraining data into account when interpreting evaluation results.
This paper proposes a data representation framework for semantic parsing and task-oriented dialogue systems, aiming to achieve a uniform representation for syntactically and semantically diverse machine-readable formats.Current NLP systems heavily rely on adapting pre-trained language models to specific tasks, and this approach has been proven effective for modeling natural language texts.However, little attention has been paid to the representation of machine-readable formats, such as database queries and dialogue states.We present a method for converting original machine-readable formats of semantic parsing and task-oriented dialogue datasets into a syntactically and semantically uniform representation.We define a meta grammar for syntactically uniform representations and translate semantically equivalent functions into a uniform vocabulary.Empirical experiments on 13 datasets show that accuracy consistently improves over original formats, revealing the advantage of the proposed representation.Additionally, we show that the proposed representation allows for transfer learning across datasets.
Entity linking faces significant challenges such as prolific variations and prevalent ambiguities, especially in high-value domains with myriad entities. Standard classification approaches suffer from the annotation bottleneck and cannot effectively handle unseen entities. Zero-shot entity linking has emerged as a promising direction for generalizing to new entities, but it still requires example gold entity mentions during training and canonical descriptions for all entities, both of which are rarely available outside of Wikipedia. In this paper, we explore Knowledge-RIch Self-Supervision (KRISS) for biomedical entity linking, by leveraging readily available domain knowledge. In training, it generates self-supervised mention examples on unlabeled text using a domain ontology and trains a contextual encoder using contrastive learning. For inference, it samples self-supervised mentions as prototypes for each entity and conducts linking by mapping the test mention to the most similar prototype. Our approach can easily incorporate entity descriptions and gold mention labels if available. We conducted extensive experiments on seven standard datasets spanning biomedical literature and clinical notes. Without using any labeled information, our method produces KRISSBERT, a universal entity linker for four million UMLS entities that attains new state of the art, outperforming prior self-supervised methods by as much as 20 absolute points in accuracy. We released KRISSBERT at https://aka.ms/krissbert.
Text-to-Image Synthesis (TIS) is a popular task to convert natural language texts into realistic images. Recently, transformer-based TIS models (such as DALL-E) have been proposed using the encoder-decoder architectures. Yet, these billion-scale TIS models are difficult to tune and deploy in resource-constrained environments. In addition, there is a lack of language-specific TIS benchmarks for Chinese, together with high-performing models with moderate sizes. In this work, we present ARTIST, A tRansformer-based Chinese Text-to-Image SynThesizer for high-resolution image generation. In ARTIST, the rich linguistic and relational knowledge facts are injected into the model to ensure better model performance without the usage of ultra-large models. We further establish a large-scale Chinese TIS benchmark with the re-production results of state-of-the-art transformer-based TIS models.Results show ARTIST outperforms previous approaches.
Chinese Grammatical Error Correction (CGEC) aims to generate a correct sentence from an erroneous sequence, where different kinds of errors are mixed. This paper divides the CGEC task into two steps, namely spelling error correction and grammatical error correction. We firstly propose a novel zero-shot approach for spelling error correction, which is simple but effective, obtaining a high precision to avoid error accumulation of the pipeline structure. To handle grammatical error correction, we design part-of-speech (POS) features and semantic class features to enhance the neural network model, and propose an auxiliary task to predict the POS sequence of the target sentence. Our proposed framework achieves a 42.11 F-0.5 score on CGEC dataset without using any synthetic data or data augmentation methods, which outperforms the previous state-of-the-art by a wide margin of 1.30 points. Moreover, our model produces meaningful POS representations that capture different POS words and convey reasonable POS transition rules.
We examine LMs’ competence of directional predicate entailments by supervised fine-tuning with prompts. Our analysis shows that contrary to their apparent success on standard NLI, LMs show limited ability to learn such directional inference; moreover, existing datasets fail to test directionality, and/or are infested by artefacts that can be learnt as proxy for entailments, yielding over-optimistic results. In response, we present BoOQA (Boolean Open QA), a robust multi-lingual evaluation benchmark for directional predicate entailments, extrinsic to existing training sets. On BoOQA, we establish baselines and show evidence of existing LM-prompting models being incompetent directional entailment learners, in contrast to entailment graphs, however limited by sparsity.
Expressing empathy is important in everyday conversations, and exploring how empathy arises is crucial in automatic response generation. Most previous approaches consider only a single factor that affects empathy. However, in practice, empathy generation and expression is a very complex and dynamic psychological process. A listener needs to find out events which cause a speaker’s emotions (emotion cause extraction), project the events into some experience (knowledge extension), and express empathy in the most appropriate way (communication mechanism).To this end, we propose a novel approach, which integrates the three components - emotion cause, knowledge graph, and communication mechanism for empathetic response generation.Experimental results on the benchmark dataset demonstrate the effectiveness of our method and show that incorporating the key components generates more informative and empathetic responses.
Response diversity has become an important criterion for evaluating the quality of open-domain dialogue generation models. However, current evaluation metrics for response diversity often fail to capture the semantic diversity of generated responses, as they mainly consider lexical aspects of the generated responses. In this paper, we introduce a new automatic evaluation metric to measure the semantic diversity of generated responses. Through human evaluation, we demonstrate that our proposed metric captures human judgments on response diversity better than existing lexical-level diversity metrics. Furthermore, motivated by analyzing an existing dialogue dataset, we propose a simple yet effective learning method that improves the semantic diversity of generated responses. Our learning method weights training samples based on the semantic distribution of the training set.We show that our learning method improves response diversity and coherency better than other baseline methods through automatic and human evaluation.
Visual question answering (VQA) is a hallmark of vision and language reasoningand a challenging task under the zero-shot setting.We propose Plug-and-Play VQA (PNP-VQA),a modular framework for zero-shot VQA.In contrast to most existing works, which require substantial adaptation of pretrained language models (PLMs) for the vision modality,PNP-VQA requires no additional training of the PLMs.Instead, we propose to use natural language and network interpretation as an intermediate representation that glues pretrained models together. We first generate question-guided informative image captions,and pass the captions to a PLM as context for question answering.Surpassing end-to-end trained baselines, PNP-VQA achieves state-of-the-art results on zero-shot VQAv2 and GQA. With 11B parameters, it outperforms the 80B-parameter Flamingo model by 8.5% on VQAv2. With 738M PLM parameters, PNP-VQA achieves an improvement of 9.1% on GQA over FewVLM with 740M PLM parameters.
Without training on labeled task data, unsupervised commonsense question answering seems challenging since it requires commonsense knowledge beyond the context of questions. Previous methods typically retrieved from traditional knowledge bases or used pre-trained language models (PrLMs) to generate fixed types of knowledge, which have poor generalization ability.In this paper, we aim to address the above limitation by leveraging the implicit knowledge stored in PrLMs and propose a two-stage prompt-based unsupervised commonsense question answering framework (TSGP). We first use knowledge generation prompts to generate the knowledge required for questions with unlimited types and possible candidate answers independent of specified choices. Then, we further utilize answer generation prompts to generate possible candidate answers independent of specified choices. Experimental results and analysis on three different commonsense reasoning tasks, CommonsenseQA, OpenBookQA, and SocialIQA, demonstrate that TSGP significantly improves the reasoning ability of language models in unsupervised settings.
Subword-level models have been the dominant paradigm in NLP. However, character-level models have the benefit of seeing each character individually, providing the model with more detailed information that ultimately could lead to better models. Recent works have shown character-level models to be competitive with subword models, but costly in terms of time and computation. Character-level models with a downsampling component alleviate this, but at the cost of quality, particularly for machine translation. This work analyzes the problems of previous downsampling methods and introduces a novel downsampling method which is informed by subwords.This new downsampling method not only outperforms existing downsampling methods, showing that downsampling characters can be done without sacrificing quality, but also leads to promising performance compared to subword models for translation.
Recent years have seen a paradigm shift in NLP towards using pretrained language models (PLM) for a wide range of tasks. However, there are many difficult design decisions to represent structures (e.g. tagged text, coreference chains) in a way such that they can be captured by PLMs. Prior work on structured prediction with PLMs typically flattens the structured output into a sequence, which limits the quality of structural information being learned and leads to inferior performance compared to classic discriminative models. In this work, we describe an approach to model structures as sequences of actions in an autoregressive manner with PLMs, allowing in-structure dependencies to be learned without any loss. Our approach achieves the new state-of-the-art on all the structured prediction tasks we looked at, namely, named entity recognition, end-to-end relation extraction, and coreference resolution.
The surge of pre-training has witnessed the rapid development of document understanding recently. Pre-training and fine-tuning framework has been effectively used to tackle texts in various formats, including plain texts, document texts, and web texts. Despite achieving promising performance, existing pre-trained models usually target one specific document format at one time, making it difficult to combine knowledge from multiple document formats. To address this, we propose XDoc, a unified pre-trained model which deals with different document formats in a single model. For parameter efficiency, we share backbone parameters for different formats such as the word embedding layer and the Transformer layers. Meanwhile, we introduce adaptive layers with lightweight parameters to enhance the distinction across different formats. Experimental results have demonstrated that with only 36.7% parameters, XDoc achieves comparable or even better performance on a variety of downstream tasks compared with the individual pre-trained models, which is cost effective for real-world deployment. The code and pre-trained models are publicly available at https://aka.ms/xdoc.
We investigate the dynamics of increasing the number of model parameters versus the number of labeled examples across a wide variety of tasks. Our exploration reveals that while scaling parameters consistently yields performance improvements, the contribution of additional examples highly depends on the task’s format. Specifically, in open question answering tasks, enlarging the training set does not improve performance. In contrast, classification, extractive question answering, and multiple choice tasks benefit so much from additional examples that collecting a few hundred examples is often “worth” billions of parameters. We hypothesize that unlike open question answering, which involves recalling specific information, solving strategies for tasks with a more restricted output space transfer across examples, and can therefore be learned with small amounts of labeled data.
Personalized chatbots focus on endowing the chatbots with a consistent personality to behave like real users and further act as personal assistants. Previous studies have explored generating implicit user profiles from the user’s dialogue history for building personalized chatbots. However, these studies only use the response generation loss to train the entire model, thus it is prone to suffer from the problem of data sparsity. Besides, they overemphasize the final generated response’s quality while ignoring the correlations and fusions between the user’s dialogue history, leading to rough data representations and performance degradation. To tackle these problems, we propose a self-supervised learning framework MCP for capturing better representations from users’ dialogue history for personalized chatbots. Specifically, we apply contrastive sampling methods to leverage the supervised signals hidden in user dialog history, and generate the pre-training samples for enhancing the model. We design three pre-training tasks based on three types of contrastive pairs from user dialogue history, namely response pairs, sequence augmentation pairs, and user pairs. We pre-train the utterance encoder and the history encoder towards the contrastive objectives and use these pre-trained encoders for generating user profiles while personalized response generation. Experimental results on two real-world datasets show a significant improvement in our proposed model MCP compared with the existing methods.
Expert Finding is an important task in Community Question Answering (CQA) platforms, which could help route questions to potential users to answer. The key is to learn representations of experts based on their historical answered questions accurately. In this paper, inspired by the strong text understanding ability of Pretrained Language modelings (PLMs), we propose a pre-training and fine-tuning expert finding framework. The core is that we design an expert-level pre-training paradigm, that effectively integrates expert interest and expertise simultaneously. Specifically different from the typical corpus-level pre-training, we treat each expert as the basic pre-training unit including all the historical answered question titles of the expert, which could fully indicate the expert interests for questions. Besides, we integrate the vote score information along with each answer of the expert into the pre-training phrase to model the expert ability explicitly. Finally, we propose a novel reputation-augmented Masked Language Model (MLM) pre-training strategy to capture the expert reputation information. In this way, our method could learn expert representation comprehensively, which then will be adopted and fine-tuned in the down-streaming expert-finding task. Extensive experimental results on six real-world CQA datasets demonstrate the effectiveness of our method.
To build a conversational agent that interacts fluently with humans, previous studies blend knowledge or personal profile into the pre-trained language model. However, the model that considers knowledge and persona at the same time is still limited, leading to hallucination and a passive way of using personas. We propose an effective dialogue agent that grounds external knowledge and persona simultaneously. The agent selects the proper knowledge and persona to use for generating the answers with our candidate scoring implemented with a poly-encoder. Then, our model generates the utterance with lesser hallucination and more engagingness utilizing retrieval augmented generation with knowledge-persona enhanced query. We conduct experiments on the persona-knowledge chat and achieve state-of-the-art performance in grounding and generation tasks on the automatic metrics. Moreover, we validate the answers from the models regarding hallucination and engagingness through human evaluation and qualitative results. We show our retriever’s effectiveness in extracting relevant documents compared to the other previous retrievers, along with the comparison of multiple candidate scoring methods. Code is available at https://github.com/dlawjddn803/INFO
Existing abstractive summarization systems are hampered by content hallucinations in which models generate text that is not directly inferable from the source alone. Annotations from prior work have shown that some of these hallucinations, while being ‘unfaithful’ to the source, are nonetheless factual. Our analysis in this paper suggests that these factual hallucinations occur as a result of the prevalence of factual yet unfaithful entities in summarization datasets. We find that these entities are not aberrations, but instead examples of additional world knowledge being readily used to latently connect entities and concepts – in this case connecting entities in the source document to those in the target summary. In our analysis and experiments, we demonstrate that connecting entities to an external knowledge base can lend provenance to many of these unfaithful yet factual entities, and further, this knowledge can be used to improve the factuality of summaries without simply making them more extractive.
Reinforcement learning (RL) is frequently employed in fine-tuning large language models (LMs), such as GPT-3, to penalize them for undesirable features of generated sequences, such as offensiveness, social bias, harmfulness or falsehood. The RL formulation involves treating the LM as a policy and updating it to maximise the expected value of a reward function which captures human preferences, such as non-offensiveness. In this paper, we analyze challenges associated with treating a language model as an RL policy and show how avoiding those challenges requires moving beyond the RL paradigm. We start by observing that the standard RL approach is flawed as an objective for fine-tuning LMs because it leads to distribution collapse: turning the LM into a degenerate distribution. Then, we analyze KL-regularised RL, a widely used recipe for fine-tuning LMs, which additionally constrains the fine-tuned LM to stay close to its original distribution in terms of Kullback-Leibler (KL) divergence. We show that KL-regularised RL is equivalent to variational inference: approximating a Bayesian posterior which specifies how to update a prior LM to conform with evidence provided by the reward function. We argue that this Bayesian inference view of KL-regularised RL is more insightful than the typically employed RL perspective. The Bayesian inference view explains how KL-regularised RL avoids the distribution collapse problem and offers a first-principles derivation for its objective. While this objective happens to be equivalent to RL (with a particular choice of parametric reward), there exist other objectives for fine-tuning LMs which are no longer equivalent to RL. That observation leads to a more general point: RL is not an adequate formal framework for problems such as fine-tuning language models. These problems are best viewed as Bayesian inference: approximating a pre-defined target distribution.
With the recent success of dense retrieval methods based on bi-encoders, studies have applied this approach to various interesting downstream retrieval tasks with good efficiency and in-domain effectiveness.Recently, we have also seen the presence of dense retrieval models in Math Information Retrieval (MIR) tasks,but the most effective systems remain classic retrieval methods that consider hand-crafted structure features.In this work, we try to combine the best of both worlds: a well-defined structure search method for effective formula search and efficient bi-encoder dense retrieval models to capture contextual similarities.Specifically, we have evaluated two representative bi-encoder models for token-level and passage-level dense retrieval on recent MIR tasks.Our results show that bi-encoder models are highly complementary to existing structure search methods, and we are able to advance the state-of-the-art on MIR datasets.
Math word problem solver requires both precise relation reasoning about quantities in the text and reliable generation for the diverse equation. Current sequence-to-tree or relation extraction methods regard this only from a fixed view, struggling to simultaneously handle complex semantics and diverse equations. However, human solving naturally involves two consistent reasoning views: top-down and bottom-up, just as math equations also can be expressed in multiple equivalent forms: pre-order and post-order. We propose a multi-view consistent contrastive learning for a more complete semantics-to-equation mapping. The entire process is decoupled into two independent but consistent views: top-down decomposition and bottom-up construction, and the two reasoning views are aligned in multi-granularity for consistency, enhancing global generation and precise reasoning. Experiments on multiple datasets across two languages show our approach significantly outperforms the existing baselines, especially on complex problems. We also show after consistent alignment, multi-view can absorb the merits of both views and generate more diverse results consistent with the mathematical laws.
Despite the important evolutions in few-shot and zero-shot learning techniques, domain specific applications still require expert knowledge and significant effort in annotating and labeling a large volume of unstructured textual data. To mitigate this problem, active learning, and meta-learning attempt to reach a high performance with the least amount of labeled data. In this paper, we introduce a novel approach to combine both lines of work by initializing an active learner with meta-learned parameters obtained through meta-training on tasks similar to the target task during active learning. In this approach we use the pre-trained BERT as our text-encoder and meta-learn its parameters with LEOPARD, which extends the model-agnostic meta-learning method by generating task dependent softmax weights to enable learning across tasks with different number of classes. We demonstrate the effectiveness of our method by performing active learning on five natural language understanding tasks and six datasets with five different acquisition functions. We train two different meta-initializations, and we use the pre-trained BERT base initialization as baseline. We observe that our approach performs better than the baseline at low budget, especially when closely related tasks were present during meta-learning. Moreover, our results show that better performance in the initial phase, i.e., with fewer labeled samples, leads to better performance when larger acquisition batches are used. We also perform an ablation study of the proposed method, showing that active learning with only the meta-learned weights is beneficial and adding the meta-learned learning rates and generating the softmax have negative consequences for the performance.
Inducing semantic representations directly from speech signals is a highly challenging task but has many useful applications in speech mining and spoken language understanding. This study tackles the unsupervised learning of semantic representations for spoken utterances. Through converting speech signals into hidden units generated from acoustic unit discovery, we propose WavEmbed, a multimodal sequential autoencoder that predicts hidden units from a dense representation of speech. Secondly, we also propose S-HuBERT to induce meaning through knowledge distillation, in which a sentence embedding model is first trained on hidden units and passes its knowledge to a speech encoder through contrastive learning. The best performing model achieves a moderate correlation (0.5 0.6) with human judgments, without relying on any labels or transcriptions. Furthermore, these models can also be easily extended to leverage textual transcriptions of speech to learn much better speech embeddings that are strongly correlated with human annotations. Our proposed methods are applicable to the development of purely data-driven systems for speech mining, indexing and search.
Multilingual transformer language models have recently attracted much attention from researchers and are used in cross-lingual transfer learning for many NLP tasks such as text classification and named entity recognition.However, similar methods for transfer learning from monolingual text to code-switched text have not been extensively explored mainly due to the following challenges:(1) Code-switched corpus, unlike monolingual corpus, consists of more than one language and existing methods can’t be applied efficiently,(2) Code-switched corpus is usually made of resource-rich and low-resource languages and upon using multilingual pre-trained language models, the final model might bias towards resource-rich language. In this paper, we focus on code-switched sentiment analysis where we have a labelled resource-rich language dataset and unlabelled code-switched data. We propose a framework that takes the distinction between resource-rich and low-resource language into account.Instead of training on the entire code-switched corpus at once, we create buckets based on the fraction of words in the resource-rich language and progressively train from resource-rich language dominated samples to low-resource language dominated samples. Extensive experiments across multiple language pairs demonstrate that progressive training helps low-resource language dominated samples.
Stance Detection Task (SDT) aims at identifying the stance of the sentence towards a specific target and is usually modeled as a classification problem. Backgound knowledge is often necessary for stance detection with respect to a specific target, especially when there is no target explicitly mentioned in text. This paper focuses on the knowledge stimulation for low-resource stance detection tasks. We firstly explore to formalize stance detection as a prompt based contrastive learning task. At the same time, to make prompt learning suit to stance detection, we design a template mechanism to incorporate corresponding target into instance representation. Furthermore, we propose a masked language prompt joint contrastive learning approach to stimulate the knowledge inherit from the pre-trained model. The experimental results on three benchmarks show that knowledge stimulation is effective in stance detection accompanied with our proposed mechanism.
Chinese spelling check (CSC) detects and corrects spelling errors in Chinese texts. Previous approaches have combined character-level phonetic and graphic information, ignoring the importance of segment-level information. According to our pilot study, spelling errors are always associated with incorrect word segmentation. When appropriate word boundaries are provided, CSC performance is greatly enhanced. Based on these findings, we present WSpeller, a CSC model that takes into account word segmentation. A fundamental component of WSpeller is a W-MLM, which is trained by predicting visually and phonetically similar words. Through modification of the embedding layer’s input, word segmentation information can be incorporated. Additionally, a robust module is trained to assist the W-MLM-based correction module by predicting the correct word segmentations from sentences containing spelling errors. We evaluate WSpeller on the widely used benchmark datasets SIGHAN13, SIGHAN14, and SIGHAN15. Our model is superior to state-of-the-art baselines on SIGHAN13 and SIGHAN15 and maintains equal performance on SIGHAN14.
A growing interest emerges in event extraction which aims to extract multiple events with triggers and arguments. Previous methods mitigate the problem of multiple events extraction by predicting the arguments conditioned on the event trigger and event type, assuming that these arguments belong to a single event. However, the assumption is invalid in general as there may be multiple events. Therefore, we present a unified framework called MatEE for trigger-sharing events extraction. It resolves the kernel bottleneck by effectively modeling the relations between arguments by an event matrix, where trigger-sharing events are represented by multiple cliques. We verify the proposed method on 3 widely-used benchmark datasets of event extraction. The experimental results show that it beats all the advanced competitors, significantly improving the state-of-the-art performances in event extraction.
Knowledge graph embedding (KGE) aims to learn continuous vector representations of relations and entities in knowledge graph (KG). Recently, transition-based KGE methods have become popular and achieved promising performance. However, scoring patterns like TransE are not suitable for complex scenarios where the same entity pair has different relations. Although some models attempt to employ entity-relation interaction or projection to improve entity representation for one-to-many/many-to-one/many-to-many complex relations, they still continue the traditional scoring pattern, where only a single relation vector in the relation part is used to translate the head entity to the tail entity or their variants. And recent research shows that entity representation only needs to consider entities and their interactions to achieve better performance. Thus, in this paper, we propose a novel transition-based method, TranS, for KGE. The single relation vector of the relation part in the traditional scoring pattern is replaced by the synthetic relation representation with entity-relation interactions to solve these issues. And the entity part still retains its independence through entity-entity interactions. Experiments on a large KG dataset, ogbl-wikikg2, show that our model achieves state-of-the-art results.
Recently, topic-grounded dialogue system has attracted significant attention due to its effectiveness in predicting the next topic to yield better responses via the historical context and given topic sequence. However, almost all existing topic prediction solutions focus on only the current conversation and corresponding topic sequence to predict the next conversation topic, without exploiting other topic-guided conversations which may contain relevant topic-transitions to current conversation. To address the problem, in this paper we propose a novel approach, named Sequential Global Topic Attention (SGTA) to exploit topic transition over all conversations in a subtle way for better modeling post-to-response topic-transition and guiding the response generation to the current conversation. Specifically, we introduce a latent space modeled as a Multivariate Skew-Normal distribution with hybrid kernel functions to flexibly integrate the global-level information with sequence-level information, and predict the topic based on the distribution sampling results. We also leverage a topic-aware prior-posterior approach for secondary selection of predicted topics, which is utilized to optimize the response generation task. Extensive experiments demonstrate that our model outperforms competitive baselines on prediction and generation tasks.
Data artifacts incentivize machine learning models to learn non-transferable generalizations by taking advantage of shortcuts in the data, andthere is growing evidence that data artifacts play a role for the strong results that deep learning models achieve in recent natural language processing benchmarks.In this paper, we focus on task-oriented dialogue and investigate whether popular datasets such as MultiWOZ contain such data artifacts.We found that by only keeping frequent phrases in the trainingexamples, state-of-the-art models perform similarly compared to the variant trained with full data, suggesting they exploit these spurious correlationsto solve the task. Motivated by this, we propose a contrastive learning based framework to encourage the model to ignore these cues and focus on learning generalisable patterns. We also experiment with adversarial filtering to remove easy training instances so that the model would focus on learning from the harder instances. We conduct a number of generalization experiments — e.g., cross-domain/dataset and adversarial tests — to assess the robustness of our approach and found that it works exceptionally well.
In this paper, we propose a novel SQL guided pre-training framework STAR for context-dependent text-to-SQL parsing, which leverages contextual information to enrich natural language (NL) utterance and table schema representations for text-to-SQL conversations. Concretely, we propose two novel pre-training objectives which respectively explore the context-dependent interactions of NL utterances and SQL queries within each text-to-SQL conversation: (i) schema state tracking (SST) objective that tracks and explores the schema states of context-dependent SQL queries in the form of schema-states by predicting and updating the value of each schema slot during interaction; (ii) utterance dependency tracking (UDT) objective that employs weighted contrastive learning to pull together two semantically similar NL utterances and push away the representations of semantically dissimilar NL utterances within each conversation. In addition, we construct a high-quality large-scale context-dependent text-to-SQL conversation corpus to pre-train STAR. Extensive experiments show that STAR achieves new state-of-the-art performance on two downstream benchmarks (SParC and CoSQL), significantly outperforming previous pre-training methods and ranking first on the leaderboard. We believe the release of the constructed corpus, codebase and pre-trained STAR checkpoints would push forward the research in this area.
Task-Oriented Dialogue (TOD) systems are drawing more and more attention in recent studies.Current methods focus on constructing pre-trained models or fine-tuning strategies while the evaluation of TOD is limited by a policy mismatch problem.That is, during evaluation, the user utterances are from the annotated dataset while these utterances should interact with previous responses which can have many alternatives besides annotated texts.Therefore, in this work, we propose an interactive evaluation framework for TOD. We first build a goal-oriented user simulator based on pre-trained models and then use the user simulator to interact with the dialogue system to generate dialogues.Besides, we introduce a sentence-level and a session-level score to measure the sentence fluency and session coherence in the interactive evaluation. Experimental results show that RL-based TOD systems trained by our proposed user simulator can achieve nearly 98% inform and success rates in the interactive evaluation of MultiWOZ dataset and the proposed scores measure the response quality besides the inform and success rates.We are hoping that our work will encourage simulator-based interactive evaluations in the TOD task.
The Annals of Joseon Dynasty (AJD) contain the daily records of the Kings of Joseon, the 500-year kingdom preceding the modern nation of Korea.The Annals were originally written in an archaic Korean writing system, ‘Hanja’, and were translated into Korean from 1968 to 1993.The resulting translation was however too literal and contained many archaic Korean words; thus, a new expert translation effort began in 2012. Since then, the records of only one king have been completed in a decade.In parallel, expert translators are working on English translation, also at a slow pace and produced only one king’s records in English so far.Thus, we propose H2KE, a neural machine translation model, that translates historical documents in Hanja to more easily understandable Korean and to English.Built on top of multilingual neural machine translation, H2KE learns to translate a historical document written in Hanja, from both a full dataset of outdated Korean translation and a small dataset of more recently translated contemporary Korean and English.We compare our method against two baselines:a recent model that simultaneously learns to restore and translate Hanja historical documentand a Transformer based model trained only on newly translated corpora.The experiments reveal that our method significantly outperforms the baselines in terms of BLEU scores for both contemporary Korean and English translations.We further conduct extensive human evaluation which shows that our translation is preferred over the original expert translations by both experts and non-expert Korean speakers.
Compositional image retrieval (CIR) is a challenging retrieval task, where the query is composed of a reference image and a modification text, and the target is another image reflecting the modification to the reference image. Due to the great success of the pre-trained vision-and-language model CLIP and its favorable applicability to large-scale retrieval tasks, we propose a CIR model HyCoLe-HNM with CLIP as the backbone. In HyCoLe-HNM, we follow the contrastive pre-training method of CLIP to perform cross-modal representation learning. On this basis, we propose a hybrid compositional learning mechanism, which includes both image compositional learning and text compositional learning. In hybrid compositional learning, we borrow a gated fusion mechanism from a question answering model to perform compositional fusion, and propose a heuristic negative mining method to filter negative samples. Privileged information in the form of image-related texts is utilized in cross-modal representation learning and hybrid compositional learning. Experimental results show that HyCoLe-HNM achieves state-of-the-art performance on three CIR datasets, namely FashionIQ, Fashion200K, and MIT-States.
While Transformer-based language models are generally very robust to pruning, there is the recently discovered outlier phenomenon: disabling only 48 out of 110M parameters in BERT-base drops its performance by nearly 30% on MNLI. We replicate the original evidence for the outlier phenomenon and we link it to the geometry of the embedding space. We find that in both BERT and RoBERTa the magnitude of hidden state coefficients corresponding to outlier dimensions correlate with the frequencies of encoded tokens in pre-training data, and they also contribute to the “vertical” self-attention pattern enabling the model to focus on the special tokens. This explains the drop in performance from disabling the outliers, and it suggests that to decrease anisotopicity in future models we need pre-training schemas that would better take into account the skewed token distributions.
Modal verbs (e.g., can, should or must) occur highly frequently in scientific articles. Decoding their function is not straightforward: they are often used for hedging, but they may also denote abilities and restrictions. Understanding their meaning is important for accurate information extraction from scientific text.To foster research on the usage of modals in this genre, we introduce the MIST (Modals In Scientific Text) dataset, which contains 3737 modal instances in five scientific domains annotated for their semantic, pragmatic, or rhetorical function. We systematically evaluate a set of competitive neural architectures on MIST. Transfer experiments reveal that leveraging non-scientific data is of limited benefit for modeling the distinctions in MIST. Our corpus analysis provides evidence that scientific communities differ in their usage of modal verbs, yet, classifiers trained on scientific data generalize to some extent to unseen scientific domains.
Prompt tuning is a parameter-efficient tuning (PETuning) method for utilizing pre-trained models (PTMs) that simply prepends a soft prompt to the input and only optimizes the prompt to adapt PTMs to downstream tasks. Although it is parameter- and deployment-efficient, its performance still lags behind other state-of-the-art PETuning methods. Besides, the training cost of prompt tuning is not significantly reduced due to the back-propagation through the entire model. Through empirical analyses, we shed some light on the lagging performance of prompt tuning and recognize a trade-off between the propagation distance from label signals to the inserted prompt and the influence of the prompt on model outputs. Further, we present Late Prompt Tuning (LPT) that inserts a late prompt into an intermediate layer of the PTM instead of the input layer or all layers. The late prompt is obtained by a neural prompt generator conditioned on the hidden states before the prompt insertion layer and therefore is instance-dependent. Through extensive experimental results across various tasks and PTMs, we show that LPT can achieve competitive performance to full model tuning and other PETuning methods under both full-data and few-shot scenarios while possessing faster training speed and lower memory cost.
Commonsense reasoning tasks such as commonsense knowledge graph completion and commonsense question answering require powerful representation learning. In this paper, we propose to learn commonsense knowledge representation by MICO, a Multi-alternative contrastIve learning framework on COmmonsense knowledge graphs (MICO). MICO generates the commonsense knowledge representation by contextual interaction between entity nodes and relations with multi-alternative contrastive learning. In MICO, the head and tail entities in an (h,r,t) knowledge triple are converted to two relation-aware sequence pairs (a premise and an alternative) in the form of natural language. Semantic representations generated by MICO can benefit the following two tasks by simply comparing the similarity score between the representations: 1) zero-shot commonsense question answering tasks; 2) inductive commonsense knowledge graph completion tasks. Extensive experiments show the effectiveness of our method.
Aspect category detection (ACD) aims to automatically identify user-concerned aspects from online reviews, which is of great value for evaluating the fine-grained performance of a product. The most recent solutions tackle this problem via weakly supervised methods, achieving remarkable improvement over unsupervised methods. However, a closer look at these methods reveals that the required human efforts are nontrivial and can sometimes be hard to obtain. In this study, we explore the possibility of minimizing human guidance while improving detection performance, with a deep clustering method that relies merely on the category name of each aspect and a pretrained language model (LM). The LM, combined with prompt techniques, is employed as a knowledge base to automatically generate constraints for clustering, as well as to provide a representation space to perform the clustering. Our method (1) extracts extensive keywords to expand our understanding of each aspect, (2) automatically generates instance-level and concept-level constraints for clustering, and (3) trains the clustering model with the above constraints. We demonstrate the capability of the proposed framework through extensive experiments on nine benchmark datasets. Our model not only performs noticeably better than existing unsupervised approaches but also considerably surpasses weakly supervised methods that require more human efforts.
In this work, we present a model for document-grounded response generation in dialog that is decomposed into two components according to Bayes’ theorem.One component is a traditional ungrounded response generation model and the other component models the reconstruction of the grounding document based on the dialog context and generated response.We propose different approximate decoding schemes and evaluate our approach on multiple open-domain and task-oriented document-grounded dialog datasets.Our experiments show that the model is more factual in terms of automatic factuality metrics than the baseline model.Furthermore, we outline how introducing scaling factors between the components allows for controlling the tradeoff between factuality and fluency in the model output.Finally, we compare our approach to a recently proposed method to control factuality in grounded dialog, CTRL (Rashkin et al., 2021), and show that both approaches can be combined to achieve additional improvements.
Causal transformer language models (LMs), such as GPT-3, typically require some form of positional encoding, such as positional embeddings. However, we show that LMs without any explicit positional encoding are still competitive with standard models and that this phenomenon is robust across different datasets, model sizes, and sequence lengths.Probing experiments reveal that such models acquire an implicit notion of absolute positions throughout the network, effectively compensating for the missing information.We conjecture that causal attention enables the model to infer the number of predecessors that each token can attend to, thereby approximating its absolute position.Our findings indicate that causal LMs might derive positional awareness not only from the explicit positioning mechanism but also from the effects of the causal mask.
Contrastive explanation methods go beyond transparency and address the contrastive aspect of explanations. Such explanations are emerging as an attractive option to provide actionable change to scenarios adversely impacted by classifiers’ decisions. However, their extension to textual data is under-explored and there is little investigation on their vulnerabilities and limitations. This work motivates textual counterfactuals by highlighting the social limitations of non-contrastive explainability. We also lay the ground for a novel evaluation scheme inspired by the faithfulness of explanations. Accordingly, we extend the computation of three metrics, proximity, connectedness and stability, to textual data and we benchmark two successful contrastive methods, POLYJUICE and MiCE, on our suggested metrics. Experiments on sentiment analysis data show that the connectedness of counterfactuals to their original counterparts is not obvious in both models. More interestingly, the generated contrastive texts are more attainable with POLYJUICE which highlights the significance of latent representations in counterfactual search. Finally, we perform the first semantic adversarial attack on textual recourse methods. The results demonstrate the robustness of POLYJUICE and the role that latent input representations play in robustness and reliability.
The attention mechanism is considered the backbone of the widely-used Transformer architecture. It contextualizes the input by computing input-specific attention matrices. We find that this mechanism, while powerful and elegant, is not as important as typically thought for pretrained language models. We introduce PAPA, a new probing method that replaces the input-dependent attention matrices with constant ones—the average attention weights over multiple inputs. We use PAPA to analyze several established pretrained Transformers on six downstream tasks. We find that without any input-dependent attention, all models achieve competitive performance—an average relative drop of only 8% from the probing baseline. Further, little or no performance drop is observed when replacing half of the input-dependent attention matrices with constant (input-independent) ones. Interestingly, we show that better-performing models lose more from applying our method than weaker models, suggesting that the utilization of the input-dependent attention mechanism might be a factor in their success. Our results motivate research on simpler alternatives to input-dependent attention, as well as on methods for better utilization of this mechanism in the Transformer architecture.
A number of knowledge integration (KI) methods have recently been proposed to incorporate external knowledge into pretrained language models (LMs). Even though knowledge-enhanced LMs (KELMs) outperform base LMs on knowledge-intensive tasks, the inner-workings of these KI methods are not well-understood. For instance, it is unclear which knowledge is effectively integrated into KELMs and which is not; and if such integration led to catastrophic forgetting of already learned knowledge. We show that existing model interpretation methods such as linear probes and prompts have some key limitations in answering these questions. Then, we revisit KI from an information-theoretic view and propose a new theoretically sound probe model called Graph Convolution Simulator (GCS) for KI interpretation. GCS is eventually quite simple – it uses graph attention on the corresponding knowledge graph for interpretation.We conduct various experiments to verify that GCS provides reasonable interpretation results for two well-known KELMs: ERNIE and K-Adapter. Our experiments reveal that only little knowledge is successfully integrated in these models, and simply increasing the size of the KI corpus may not lead to better KELMs.
Open Information Extraction (OpenIE) facilitates the open-domain discovery of textual facts. However, the prevailing solutions evaluate OpenIE models on in-domain test sets aside from the training corpus, which certainly violates the initial task principle of domain-independence. In this paper, we propose to advance OpenIE towards a more realistic scenario: generalizing over unseen target domains with different data distributions from the source training domains, termed Generalized OpenIE. For this purpose, we first introduce GLOBE, a large-scale human-annotated multi-domain OpenIE benchmark, to examine the robustness of recent OpenIE models to domain shifts, and the relative performance degradation of up to 70% implies the challenges of generalized OpenIE. Then, we propose DragonIE, which explores a minimalist expression of textual fact: directed acyclic graph, to improve the OpenIE generalization ability. Extensive experiments demonstrate that DragonIE beats the previous methods in both in-domain and out-of-domain settings by as much as 6.0% in F1 score absolutely, but there is still ample room for improvement.
This work introduces BioLORD, a new pre-training strategy for producing meaningful representations for clinical sentences and biomedical concepts. State-of-the-art methodologies operate by maximizing the similarity in representation of names referring to the same concept, and preventing collapse through contrastive learning. However, because biomedical names are not always self-explanatory, it sometimes results in non-semantic representations. BioLORD overcomes this issue by grounding its concept representations using definitions, as well as short descriptions derived from a multi-relational knowledge graph consisting of biomedical ontologies. Thanks to this grounding, our model produces more semantic concept representations that match more closely the hierarchical structure of ontologies. BioLORD establishes a new state of the art for text similarity on both clinical sentences (MedSTS) and biomedical concepts (MayoSRS).
In parsing phrase structures, supertagging achieves a symbiosis between the interpretability of formal grammars and the accuracy and speed of more recent neural models.The approach was only recently transferred to parsing discontinuous constituency structures with linear context-free rewriting systems (LCFRS).We reformulate and parameterize the previously fixed extraction process for LCFRS supertags with the aim to improve the overall parsing quality.These parameters are set in the context of several steps in the extraction process and are used to control the granularity of extracted grammar rules as well as the association of lexical symbols with each supertag.We evaluate the influence of the parameters on the sets of extracted supertags and the parsing quality using three treebanks in the English and German language, and we compare the best-performing configurations to recent state-of-the-art parsers in the area.Our results show that some of our configurations and the slightly modified parsing process improve the quality and speed of parsing with our supertags over the previous approach.Moreover, we achieve parsing scores that either surpass or are among the state-of-the-art in discontinuous constituent parsing.
The pre-training of masked language models (MLMs) consumes massive computation to achieve good results on downstream NLP tasks, resulting in a large carbon footprint. In the vanilla MLM, the virtual tokens, [MASK]s, act as placeholders and gather the contextualized information from unmasked tokens to restore the corrupted information. It raises the question of whether we can append [MASK]s at a later layer, to reduce the sequence length for earlier layers and make the pre-training more efficient. We show: (1) [MASK]s can indeed be appended at a later layer, being disentangled from the word embedding; (2) The gathering of contextualized information from unmasked tokens can be conducted with a few layers. By further increasing the masking rate from 15% to 50%, we can pre-train RoBERTa-base and RoBERTa-large from scratch with only 78% and 68% of the original computational budget without any degradation on the GLUE benchmark. When pre-training with the original budget, our method outperforms RoBERTa for 6 out of 8 GLUE tasks, on average by 0.4%.
Word Sense Disambiguation (WSD) is an NLP task aimed at determining the correct sense of a word in a sentence from discrete sense choices. Although current systems have attained unprecedented performances for such tasks, the nonuniform distribution of word senses during training generally results in systems performing poorly on rare senses. To this end, we consider data augmentation to increase the frequency of these least frequent senses (LFS) to reduce the distributional bias of senses during training. We propose Sense-Maintained Sentence Mixup (SMSMix), a novel word-level mixup method that maintains the sense of a target word. SMSMix smoothly blends two sentences using mask prediction while preserving the relevant span determined by saliency scores to maintain a specific word’s sense. To the best of our knowledge, this is the first attempt to apply mixup in NLP while preserving the meaning of a specific word. With extensive experiments, we validate that our augmentation method can effectively give more information about rare senses during training with maintained target sense label.
A major challenge in the field of Text Generation is evaluation, because we lack a sound theory that can be leveraged to extract guidelines for evaluation campaigns. In this work, we propose a first step towards such a theory that incorporates different sources of uncertainty, such as imperfect automated metrics and insufficiently sized test sets. The theory has practical applications, such as determining the number of samples needed to reliably distinguish the performance of a set of Text Generation systems in a given setting. We showcase the application of the theory on the WMT 21 and Spot-The-Bot evaluation data and outline how it can be leveraged to improve the evaluation protocol regarding the reliability, robustness, and significance of the evaluation outcome.
N-gram language models (LM) has been largely superseded by neural LMs as the latter exhibits better performance. However, we find that n-gram models can achieve satisfactory performance on a large proportion of testing cases, indicating they have already captured abundant knowledge of the language with relatively low computational cost. With this observation, we propose to learn a neural LM that fits the residual between an n-gram LM and the real-data distribution. The combination of n-gram LMs and neural LMs not only allows the neural part to focus on deeper understanding of the language, but also provides a flexible way to customize a LM by switching the underlying n-gram model without changing the neural model. Experimental results on three typical language tasks (i.e., language modeling, machine translation, and summarization) demonstrate that our approach attains additional performance gains over popular standalone neural models consistently. We also show that our approach allows for effective domain adaptation by simply switching to a domain-specific n-gram model, without any extra training.
Taking into account background knowledge as the context has always been an important part of solving tasks that involve natural language. One representative example of such tasks is text-based games, where players need to make decisions based on both description text previously shown in the game, and their own background knowledge about the language and common sense. In this work, we investigate not simply giving common sense, as can be seen in prior research, but also its effective usage. We assume that a part of the environment states different from common sense should constitute one of the grounds for action selection. We propose a novel agent, DiffG-RL, which constructs a Difference Graph that organizes the environment states and common sense by means of interactive objects with a dedicated graph encoder. DiffG-RL also contains a framework for extracting the appropriate amount and representation of common sense from the source to support the construction of the graph. We validate DiffG-RL in experiments with text-based games that require common sense and show that it outperforms baselines by 17% of scores. We will make our code publicly available.
Syntactically controlled paraphrase generation has become an emerging research direction in recent years. Most existing approaches require annotated paraphrase pairs for training and are thus costly to extend to new domains. Unsupervised approaches, on the other hand, do not need paraphrase pairs but suffer from relatively poor performance in terms of syntactic control and quality of generated paraphrases. In this paper, we demonstrate that leveraging Abstract Meaning Representations (AMR) can greatly improve the performance of unsupervised syntactically controlled paraphrase generation.Our proposed model, AMR-enhanced Paraphrase Generator (AMRPG), separately encodes the AMR graph and the constituency parse of the input sentence into two disentangled semantic and syntactic embeddings. A decoder is then learned to reconstruct the input sentence from the semantic and syntactic embeddings. Our experiments show that AMRPG generates more accurate syntactically controlled paraphrases, both quantitatively and qualitatively, compared to the existing unsupervised approaches. We also demonstrate that the paraphrases generated by AMRPG can be used for data augmentation to improve the robustness of NLP models.
Abstract Meaning Representation (AMR) has been shown to be useful for many downstream tasks. In this work, we explore the use of AMR for legal and logical reasoning. Specifically, we investigate if AMR can help capture logical relationships on multiple choice question answering (MCQA) tasks. We propose neural architectures that utilize linearised AMR graphs in combination with pre-trained language models. While these models are not able to outperform text-only baselines, they correctly solve different instances than the text models, suggesting complementary abilities. Error analysis further reveals that AMR parsing quality is the most prominent challenge, especially regarding inputs with multiple sentences. We conduct a theoretical analysis of how logical relations are represented in AMR and conclude it might be helpful in some logical statements but not for others.
Neural Machine Translation (NMT) models are typically trained on heterogeneous data that are concatenated and randomly shuffled. However, not all of the training data are equally useful to the model. Curriculum training aims to present the data to the NMT models in a meaningful order. In this work, we introduce a two-stage training framework for NMT where we fine-tune a base NMT model on subsets of data, selected by both deterministic scoring using pre-trained methods and online scoring that considers prediction scores of the emerging NMT model. Through comprehensive experiments on six language pairs comprising low- and high-resource languages from WMT’21, we have shown that our curriculum strategies consistently demonstrate better quality (up to +2.2 BLEU improvement) and faster convergence (approximately 50% fewer updates).
Text editing, such as grammatical error correction, arises naturally from imperfect textual data. Recent works frame text editing as a multi-round sequence tagging task, where operations – such as insertion and substitution – are represented as a sequence of tags. While achieving good results, this encoding is limited in flexibility as all actions are bound to token-level tags. In this work, we reformulate text editing as an imitation game using behavioral cloning. Specifically, we convert conventional sequence-to-sequence data into state-to-action demonstrations, where the action space can be as flexible as needed. Instead of generating the actions one at a time, we introduce a dual decoders structure to parallel the decoding while retaining the dependencies between action tokens, coupled with trajectory augmentation to alleviate the distribution shift that imitation learning often suffers. In experiments on a suite of Arithmetic Equation benchmarks, our model consistently outperforms the autoregressive baselines in terms of performance, efficiency, and robustness. We hope our findings will shed light on future studies in reinforcement learning applying sequence-level action generation to natural language processing.
Practitioners from many disciplines (e.g., political science) use expert-crafted taxonomies to make sense of large, unlabeled corpora. In this work, we study Seeded Hierarchical Clustering (SHC): the task of automatically fitting unlabeled data to such taxonomies using a small set of labeled examples. We propose HierSeed, a novel weakly supervised algorithm for this task that uses only a small set of labeled seed examples in a computation and data efficient manner. HierSeed assigns documents to topics by weighing document density against topic hierarchical structure. It outperforms unsupervised and supervised baselines for the SHC task on three real-world datasets.
In this work we propose a novel end-to-end multi-stage Knowledge Graph (KG) generation system from textual inputs, separating the overall process into two stages. The graph nodes are generated first using pretrained language model, followed by a simple edge construction head, enabling efficient KG extraction from the text. For each stage we consider several architectural choices that can be used depending on the available training resources. We evaluated the model on a recent WebNLG 2020 Challenge dataset, matching the state-of-the-art performance on text-to-RDF generation task, as well as on New York Times (NYT) and a large-scale TekGen datasets, showing strong overall performance, outperforming the existing baselines. We believe that the proposed system can serve as a viable KG construction alternative to the existing linearization or sampling-based graph generation approaches.
Financial risk prediction is an essential task for risk management in capital markets. While traditional prediction models are built based on the hard information of numerical data, recent studies have shown that the soft information of verbal cues in earnings conference calls is significant for predicting market risk due to its less constrained fashion and direct interaction between managers and analysts. However, most existing models mainly focus on extracting useful semantic information from the textual conference call transcripts but ignore their subtle yet important information of dialogue structures. To bridge this gap, we develop a graph attention network called DialogueGAT for financial risk prediction by simultaneously modeling the speakers and their utterances in dialogues in conference calls. Different from previous studies, we propose a new method for constructing the graph of speakers and utterances in a dialogue, and design contextual attention at both speaker and utterance levels for disentangling their effects on the downstream prediction task. For model evaluation, we extend an existing dataset of conference call transcripts by adding the dialogue structure and speaker information. Empirical results on our dataset of S&P1500 companies demonstrate the superiority of our proposed model over competitive baselines from the extant literature.
Large pre-trained language models have shown remarkable performance over the past few years. These models, however, sometimes learn superficial features from the dataset and cannot generalize to the distributions that are dissimilar to the training scenario. There have been several approaches proposed to reduce model’s reliance on these bias features which can improve model robustness in the out-of-distribution setting. However, existing methods usually use a fixed low-capacity model to deal with various bias features, which ignore the learnability of those features. In this paper, we analyze a set of existing bias features and demonstrate there is no single model that works best for all the cases. We further show that by choosing an appropriate bias model, we can obtain a better robustness result than baselines with a more sophisticated model design.
To obtain a large amount of training labels inexpensively, researchers have recently adopted the weak supervision (WS) paradigm, which leverages labeling rules to synthesize training labels rather than using individual annotations to achieve competitive results for natural language processing (NLP) tasks. However, data imbalance is often overlooked in applying the WS paradigm, despite being a common issue in a variety of NLP tasks. To address this challenge, we propose Adaptive Ranking-based Sample Selection (ARS2), a model-agnostic framework to alleviate the data imbalance issue in the WS paradigm. Specifically, it calculates a probabilistic margin score based on the output of the current model to measure and rank the cleanliness of each data point. Then, the ranked data are sampled based on both class-wise and rule-aware ranking. In particular, the two sample strategies corresponds to our motivations: (1) to train the model with balanced data batches to reduce the data imbalance issue and (2) to exploit the expertise of each labeling rule for collecting clean samples. Experiments on four text classification datasets with four different imbalance ratios show that ARS2 outperformed the state-of-the-art imbalanced learning and WS methods, leading to a 2%-57.8% improvement on their F1-score.
Understanding rich narratives, such as dialogues and stories, often requires natural language processing systems to access relevant knowledge from commonsense knowledge graphs. However, these systems typically retrieve facts from KGs using simple heuristics that disregard the complex challenges of identifying situationally-relevant commonsense knowledge (e.g., contextualization, implicitness, ambiguity).In this work, we propose the new task of commonsense fact linking, where models are given contexts and trained to identify situationally-relevant commonsense knowledge from KGs. Our novel benchmark, ComFact, contains ~293k in-context relevance annotations for commonsense triplets across four stylistically diverse dialogue and storytelling datasets. Experimental results confirm that heuristic fact linking approaches are imprecise knowledge extractors. Learned fact linking models demonstrate across-the-board performance improvements (~34.6% F1) over these heuristics. Furthermore, improved knowledge retrieval yielded average downstream improvements of 9.8% for a dialogue response generation task. However, fact linking models still significantly underperform humans, suggesting our benchmark is a promising testbed for research in commonsense augmentation of NLP systems.
How to usefully encode compositional task structure has long been a core challenge in AI. Recent work in chain of thought prompting has shown that for very large neural language models (LMs), explicitly demonstrating the inferential steps involved in a target task may improve performance over end-to-end learning that focuses on the target task alone. However, chain of thought prompting has significant limitations due to its dependency on huge pretrained LMs. In this work, we present compositional fine-tuning (CFT): an approach based on explicitly decomposing a target task into component tasks, and then fine-tuning smaller LMs on a curriculum of such component tasks. We apply CFT to recommendation tasks in two domains, world travel and local dining, as well as a previously studied inferential task (sports understanding). We show that CFT outperforms end-to-end learning even with equal amounts of data, and gets consistently better as more component tasks are modeled via fine-tuning. Compared with chain of thought prompting, CFT performs at least as well using LMs only 7.4% of the size, and is moreover applicable to task domains for which data are not available during pretraining.
Topic taxonomies display hierarchical topic structures of a text corpus and provide topical knowledge to enhance various NLP applications. To dynamically incorporate new topic information, several recent studies have tried to expand (or complete) a topic taxonomy by inserting emerging topics identified in a set of new documents. However, existing methods focus only on frequent terms in documents and the local topic-subtopic relations in a taxonomy, which leads to limited topic term coverage and fails to model the global taxonomy structure. In this work, we propose a novel framework for topic taxonomy expansion, named TopicExpan, which directly generates topic-related terms belonging to new topics. Specifically, TopicExpan leverages the hierarchical relation structure surrounding a new topic and the textual content of an input document for topic term generation. This approach encourages newly-inserted topics to further cover important but less frequent terms as well as to keep their relation consistency within the taxonomy. Experimental results on two real-world text corpora show that TopicExpan significantly outperforms other baseline methods in terms of the quality of output taxonomies.
The way we talk carries information about who we are. Demographics, personality, clinical conditions, political preferences influence what we speak about and how, suggesting that many individual attributes could be inferred from adequate encodings of linguistic behavior. Conversely, conditioning text representations on author attributes has been shown to improve model performance in many NLP tasks. Previous research on individual differences and language representations has mainly focused on predicting selected attributes from text, or on conditioning text representations on such attributes for author-based contextualization. Here, we present a self-supervised approach to learning language-based user encodings using transformers. Using a large corpus of Reddit submissions, we fine-tune DistilBERT on user-based triplet loss. We show that fine-tuned models can pick up on complex linguistic signatures of users, and that they are able to infer rich information about them. Through a series of intrinsic analyses and probing tasks, we provide evidence that fine-tuning enhances models’ ability to abstract generalizable user information, which yields performance advantages for user-based downstream tasks. We discuss applications in language-based assessment and contextualized and personalized NLP.
We investigate input-conditioned hypernetworks for multi-tasking in NLP, generating parameter-efficient adaptations for a decoder using a hypernetwork conditioned on the output of an encoder. This approach produces a unique decoder adaptation for every input instance, allowing the network a larger degree of flexibility than prior work that only produces one decoder adaptation per task. We apply our method to sequence classification tasks, extractive QA, and summarisation and find that it surpasses previous parameter efficient fine-tuning methods and often outperforms fully finetuning the underlying model. An analysis of the embeddings used by our hypernetwork shows that they are sensitive to output label and type, suggesting that our approach better maps from encoder representations to output labels. Our code is publicly available at https://github.com/allenai/hyperdecoders.
To explain NLP models a popular approach is to use importance measures, such as attention, which inform input tokens are important for making a prediction. However, an open question is how well these explanations accurately reflect a model’s logic, a property called faithfulness. To answer this question, we propose Recursive ROAR, a new faithfulness metric. This works by recursively masking allegedly important tokens and then retraining the model. The principle is that this should result in worse model performance compared to masking random tokens. The result is a performance curve given a masking-ratio. Furthermore, we propose a summarizing metric using area-between-curves (ABC), which allows for easy comparison across papers, models, and tasks. We evaluate 4 different importance measures on 8 different datasets, using both LSTM-attention models and RoBERTa models. We find that the faithfulness of importance measures is both model-dependent and task-dependent. This conclusion contradicts previous evaluations in both computer vision and faithfulness of attention literature.
Large-scale language models have been reducing the gap between machines and humans in understanding the real world, yet understanding an individual’s theory of mind and behavior from text is far from being resolved. This research proposes a neural model—Subjective Ground Attention—that learns subjective grounds of individuals and accounts for their judgments on situations of others posted on social media. Using simple attention modules as well as taking one’s previous activities into consideration, we empirically show that our model provides human-readable explanations of an individual’s subjective preference in judging social situations. We further qualitatively evaluate the explanations generated by the model and claim that our model learns an individual’s subjective orientation towards abstract moral concepts.
Automatic International Classification of Diseases (ICD) coding aims to assign multiple ICD codes to a medical note with average length of 3,000+ tokens. This task is challenging due to a high-dimensional space of multi-label assignment (tens of thousands of ICD codes) and the long-tail challenge: only a few codes (common diseases) are frequently assigned while most codes (rare diseases) are infrequently assigned. This study addresses the long-tail challenge by adapting a prompt-based fine-tuning technique with label semantics, which has been shown to be effective under few-shot setting. To further enhance the performance in medical domain, we propose a knowledge-enhanced longformer by injecting three domain-specific knowledge: hierarchy, synonym, and abbreviation with additional pretraining using contrastive learning. Experiments on MIMIC-III-full, a benchmark dataset of code assignment, show that our proposed method outperforms previous state-of-the-art method in 14.5% in marco F1 (from 10.3 to 11.8, P<0.001). To further test our model on few-shot setting, we created a new rare diseases coding dataset, MIMIC-III-rare50, on which our model improves marco F1 from 17.1 to 30.4 and micro F1 from 17.2 to 32.6 compared to previous method.
Recent success of pre-trained language models (PLMs) has stimulated interest in their ability to understand and work with numbers. Yet, the numerical reasoning over measurements has not been formally studied despite their importance. In this study, we show that PLMs lack the capability required for reasoning over measurements. Furthermore, we find that a language model trained on a measurement-rich corpus shows better performance on understanding measurements. We propose a simple embedding strategy to better distinguish between numbers and units, which leads to a significant improvement in the probing tasks.
Incorporating large-scale pre-trained models with the prototypical neural networks is a de-facto paradigm in few-shot named entity recognition. Existing methods, unfortunately, are not aware of the fact that embeddings from pre-trained models contain a prominently large amount of information regarding word frequencies, biasing prototypical neural networks against learning word entities. This discrepancy constrains the two models’ synergy. Thus, we propose a one-line-code normalization method to reconcile such a mismatch with empirical and theoretical grounds. Our experiments based on nine benchmark datasets show the superiority of our method over the counterpart models and are comparable to the state-of-the-art methods. In addition to the model enhancement, our work also provides an analytical viewpoint for addressing the general problems in few-shot name entity recognition or other tasks that rely on pre-trained models or prototypical neural networks.
Event detection has been suffering from constantly emerging event types with lack of sufficient data. Existing works formulate the new problem as few-shot event detection (FSED), and employ two-stage or unified models based on meta-learning to address the problem. However, these methods fall far short of expectations due to: (i) insufficient learning of discriminative representations in low-resource scenarios, and (ii) representation overlap between triggers and non-triggers. To resolve the above issues, in this paper, we propose a novel Hybrid Contrastive Learning method with a Task-Adaptive Threshold (abbreviated as HCL-TAT), which enables discriminative representation learning with a two-view contrastive loss (support-support and prototype-query), and devises an easily-adapted threshold to alleviate misidentification of triggers. Extensive experiments on the benchmark dataset FewEvent demonstrate the superiority of our method to achieve better results compared to the state-of-the-arts. All the data and codes will be available to facilitate future research.
This paper introduces Doc2Bot, a novel dataset for building machines that help users seek information via conversations. This is of particular interest for companies and organizations that own a large number of manuals or instruction books. Despite its potential, the nature of our task poses several challenges: (1) documents contain various structures that hinder the ability of machines to comprehend, and (2) user information needs are often underspecified. Compared to prior datasets that either focus on a single structural type or overlook the role of questioning to uncover user needs, the Doc2Bot dataset is developed to target such challenges systematically. Our dataset contains over 100,000 turns based on Chinese documents from five domains, larger than any prior document-grounded dialog dataset for information seeking. We propose three tasks in Doc2Bot: (1) dialog state tracking to track user intentions, (2) dialog policy learning to plan system actions and contents, and (3) response generation which generates responses based on the outputs of the dialog policy. Baseline methods based on the latest deep learning models are presented, indicating that our proposed tasks are challenging and worthy of further research.
We present DualNER, a simple and effective framework to make full use of both annotated source language corpus and unlabeled target language text for zero-shot cross-lingual named entity recognition (NER). In particular, we combine two complementary learning paradigms of NER, i.e., sequence labeling and span prediction, into a unified multi-task framework. After obtaining a sufficient NER model trained on the source data, we further train it on the target data in a dual-teaching manner, in which the pseudo-labels for one task are constructed from the prediction of the other task. Moreover, based on the span prediction, an entity-aware regularization is proposed to enhance the intrinsic cross-lingual alignment between the same entities in different languages. Experiments and analysis demonstrate the effectiveness of our DualNER.
The recent rise of conversational applications such as online customer service systems and intelligent personal assistants has promoted the development of conversational knowledge base question answering (ConvKBQA). Different from the traditional single-turn KBQA, ConvKBQA usually explores multi-turn questions around a topic, where ellipsis and coreference pose great challenges to the single-turn KBQA systems which require self-contained questions. In this paper, we propose a rewrite-and-reason framework to first produce a full-fledged rewritten question based on the conversation history and then reason the answer by existing single-turn KBQA models. To overcome the absence of the rewritten supervision signals, we introduce a knowledge-augmented self-training mechanism to transfer the question rewriter from another dataset to adapt to the current knowledge base. Our question rewriter is decoupled from the subsequent QA process, which makes it easy to be united with either retrieval-based or semantic parsing-based KBQA models. Experiment results demonstrate the effectiveness of our method and a new state-of-the-art result is achieved. The code and dataset are available online now.
Summarizing legal decisions requires the expertise of law practitioners, which is both time- and cost-intensive. This paper presents techniques for extractive summarization of legal decisions in a low-resource setting using limited expert annotated data. We test a set of models that locate relevant content using a sequential model and tackle redundancy by leveraging maximal marginal relevance to compose summaries. We also demonstrate an implicit approach to help train our proposed models generate more informative summaries. Our multi-task learning model variant leverages rhetorical role identification as an auxiliary task to further improve the summarizer. We perform extensive experiments on datasets containing legal decisions from the US Board of Veterans’ Appeals and conduct quantitative and expert-ranked evaluations of our models. Our results show that the proposed approaches can achieve ROUGE scores vis-à-vis expert extracted summaries that match those achieved by inter-annotator comparison.
Automatic movie narration generation and narration grounding are very important to provide a true movie experience for the blind and visually impaired. To tell the movie story well, it is necessary to mention plot-related details (such as character names) and keep the narrations in a plot coherent. Taking these two points into consideration, we construct a Chinese large-scale video benchmark from 101 movies for Movie Understanding and Narrating (MovieUN) to support the Movie Clip Narrating (MCN) task and Temporal Narration Grounding (TNG) task. We split movies in MovieUN into movie clips according to plots, and pair them with corresponding narrations provided by the movie narrators. Ultimately, the TNG task involves 3,253 long video clips totaling 179 hours. The MCN task contains 33,060 video clips totaling 105 hours. We benchmark state-of-the-art video captioning models and temporal grounding models in MCN and TNG tasks, respectively. Furthermore, to accurately comprehend plots of different characters, we propose methods to incorporate portraits of actors as external knowledge in both tasks. The experiment results demonstrate the effectiveness of our proposed methods. The dataset and codes are released at https://github.com/yuezih/MovieUN.
Data-to-text generation is challenging due to the great variety of the input data in terms of domains (e.g., finance vs sports) or schemata (e.g., diverse predicates). Recent end-to-end neural methods thus require substantial training examples to learn to disambiguate and describe the data. Yet, real-world data-to-text problems often suffer from various data-scarce issues: one may have access to only a handful of or no training examples, and/or have to rely on examples in a different domain or schema. To fill this gap, we propose Any-Shot Data-to-Text (ASDOT), a new approach flexibly applicable to diverse settings by making efficient use of any given (or no) examples. ASDOT consists of two steps, data disambiguation and sentence fusion, both of which are amenable to be solved with off-the-shelf pretrained language models (LMs) with optional finetuning. In the data disambiguation stage, we employ the prompted GPT-3 model to understand possibly ambiguous triples from the input data and convert each into a short sentence with reduced ambiguity. The sentence fusion stage then uses an LM like T5 to fuse all the resulting sentences into a coherent paragraph as the final description. We evaluate extensively on various datasets in different scenarios, including the zero-/few-/full-shot settings, and generalization to unseen predicates and out-of-domain data. Experimental results show that ASDOT consistently achieves significant improvement over baselines, e.g., a 30.81 BLEU gain on the DART dataset under the zero-shot setting.
Grammatical Error Correction (GEC) has been broadly applied in automatic correction and proofreading system recently. However, it is still immature in Chinese GEC due to limited high-quality data from native speakers in terms of category and scale. In this paper, we present FCGEC, a fine-grained corpus to detect, identify and correct the grammatical errors. FCGEC is a human-annotated corpus with multiple references, consisting of 41,340 sentences collected mainly from multi-choice questions in public school Chinese examinations. Furthermore, we propose a Switch-Tagger-Generator (STG) baseline model to correct the grammatical errors in low-resource settings. Compared to other GEC benchmark models, experimental results illustrate that STG outperforms them on our FCGEC. However, there exists a significant gap between benchmark models and humans that encourages future models to bridge it.
Adopting contextually appropriate, audience-tailored linguistic styles is critical to the success of user-centric language generation systems (e.g., chatbots, computer-aided writing, dialog systems). While existing approaches demonstrate text style transfer (TST) with large volumes of parallel or non-parallel data, we argue that grounding style on audience-independent external factors is innately limiting for two reasons. First, it is difficult to collect large volumes of audience-specific stylistic data. Second, some stylistic objectives (e.g., persuasiveness, memorability, empathy) are hard to define without audience feedback. In this paper, we propose the novel task of style infusion - infusing the stylistic preferences of audiences in pretrained language generation models. Since humans are better at pairwise comparisons than direct scoring - i.e., is Sample-A more persuasive/polite/empathic than Sample-B - we leverage limited pairwise human judgments to bootstrap a style analysis model and augment our seed set of judgments. We then infuse the learned textual style in a GPT-2 based text generator while balancing fluency and style adoption. With quantitative and qualitative assessments, we show that our infusion approach can generate compelling stylized examples with generic text prompts. We make the anonymized code and data accessible.
Financial prediction is complex due to the stochastic nature of the stock market. Semi-structured financial documents present comprehensive financial data in tabular formats, such as earnings, profit-loss statements, and balance sheets, and can often contain rich technical analysis along with a textual discussion of corporate history, and management analysis, compliance, and risks. Existing research focuses on the textual and audio modalities of financial disclosures from company conference calls to forecast stock volatility and price movement, but ignores the rich tabular data available in financial reports. Moreover, the economic realm is still plagued with a severe under-representation of various communities spanning diverse demographics, gender, and native speakers. In this work, we show that combining tabular data from financial semi-structured documents with text transcripts and audio recordings not only improves stock volatility and price movement prediction by 5-12% but also reduces gender bias caused due to audio-based neural networks by over 30%.
Document-level relation extraction (DocRE) aims to identify semantic labels among entities within a single document. One major challenge of DocRE is to dig decisive details regarding a specific entity pair from long text. However, in many cases, only a fraction of text carries required information, even in the manually labeled supporting evidence. To better capture and exploit instructive information, we propose a novel expLicit syntAx Refinement and Subsentence mOdeliNg based framework (LARSON). By introducing extra syntactic information, LARSON can model subsentences of arbitrary granularity and efficiently screen instructive ones. Moreover, we incorporate refined syntax into text representations which further improves the performance of LARSON. Experimental results on three benchmark datasets (DocRED, CDR, and GDA) demonstrate that LARSON significantly outperforms existing methods.
Pre-trained speech Transformers have facilitated great success across various speech processing tasks. However, fine-tuning these encoders for downstream tasks require sufficiently large training data to converge or to achieve state-of-the-art. In text domain this has been partly attributed to sub-optimality of the representation space in pre-trained Transformers. In this work, we take a sober look into pre-trained speech encoders and rewire their representation space without requiring any task-specific labels. Our method utilises neutrally synthesised version of audio inputs along with frame masking to construct positive pairs for contrastive self-supervised learning. When used for augmenting the wav2vec 2 encoder, we observe consistent improvement of isotropy in the representation space. Our experiments on 6 speech processing tasks, exhibit a significant convergence speedup during task fine-tuning as well as consistent task improvement, specially in low-resource settings.
Pre-trained speech Transformers in speech translation (ST) have facilitated state-of-the-art (SotA) results; yet, using such encoders is computationally expensive. To improve this, we present a novel Reducer Adaptor block, RedApt, that could be seamlessly integrated within any Transformer-based speech encoding architecture. Integrating the pretrained wav2vec 2 speech encoder with RedAptbrings 41% speedup, 33% memory reduction with 24% fewer FLOPs at inference. To our positive surprise, our ST model with RedApt outperforms the SotA architecture by an average of 0.68 BLEU score on 8 language pairs from Must-C.
Neural Machine Translation systems built on top of Transformer-based architectures are routinely improving the state-of-the-art in translation quality according to word-overlap metrics. However, a growing number of studies also highlight the inherent gender bias that these models incorporate during training, which reflects poorly in their translations. In this work, we investigate whether these models can be instructed to fix their bias during inference using targeted, guided instructions as contexts. By translating relevant contextual sentences during inference along with the input, we observe large improvements in reducing the gender bias in translations, across three popular test suites (WinoMT, BUG, SimpleGen). We further propose a novel metric to assess several large pre-trained models (OPUS-MT, M2M-100) on their sensitivity towards using contexts during translation to correct their biases. Our approach requires no fine-tuning, and thus can be used easily in production systems to de-bias translations from stereotypical gender-occupation bias. We hope our method, along with our metric, can be used to build better, bias-free translation systems.
Clinical outcome prediction is critical to the condition prediction of patients and management of hospital capacities. There are two kinds of medical data, including time series signals recorded by various devices and clinical notes in electronic health records (EHR), which are used for two common prediction targets: mortality and length of stay. Traditional methods focused on utilizing time series data but ignored clinical notes. With the development of deep learning, natural language processing (NLP) and multi-modal learning methods are exploited to jointly model the time series and clinical notes with different modals. However, the existing methods failed to fuse the multi-modal features of patients from different views. Therefore, we propose the patient multi-view multi-modal feature fusion networks for clinical outcome prediction. Firstly, from patient inner view, we propose to utilize the co-attention module to enhance the fine-grained feature interaction between time series and clinical notes from each patient. Secondly, the patient outer view is the correlation between patients, which can be reflected by the structural knowledge in clinical notes. We exploit the structural information extracted from clinical notes to construct the patient correlation graph, and fuse patients’ multi-modal features by graph neural networks (GNN). The experimental results on MIMIC-III benchmark demonstrate the superiority of our method.
Automatic document summarization aims to produce a concise summary covering the input document’s salient information. Within a report document, the salient information can be scattered in the textual and non-textual content. However, existing document summarization datasets and methods usually focus on the text and filter out the non-textual content. Missing tabular data can limit produced summaries’ informativeness, especially when summaries require covering quantitative descriptions of critical metrics in tables. Existing datasets and methods cannot meet the requirements of summarizing long text and multiple tables in each report. To deal with the scarcity of available data, we propose FINDSum, the first large-scale dataset for long text and multi-table summarization. Built on 21,125 annual reports from 3,794 companies, it has two subsets for summarizing each company’s results of operations and liquidity. To summarize the long text and dozens of tables in each report, we present three types of summarization methods. Besides, we propose a set of evaluation metrics to assess the usage of numerical information in produced summaries. Dataset analyses and experimental results indicate the importance of jointly considering input textual and tabular data when summarizing report documents.
Text ranking plays a key role in providing content that best answers user queries. It is usually divided into two sub-tasks to perform efficient information retrieval given a query: text retrieval and text re-ranking. Recent research on pretrained language models (PLM) has demonstrated efficiency and gain on both sub-tasks. However, while existing methods have benefited from pre-trained language models and achieved high recall rates on passage retrieval, the ranking performance still demands further improvement. In this paper, we propose MatRank, which learns to re-rank the text retrieved for a given query by learning to predict the most relevant passage based on a latent preference matrix. Specifically, MatRank uses a PLM to generate an asymmetric latent matrix of relative preference scores between all pairs of retrieved passages. Then, the latent matrix is aggregated row-wise and column-wise to obtain global preferences and predictions of the most relevant passage in two of these directions, respectively. We conduct extensive experiments on MS MACRO, WikiAQ, and SemEval datasets. Experimental results show that MatRank has achieved new state-of-the-art results on these datasets, outperforming all prior methods on ranking performance metrics.
Recent progress regarding the use of language models (LMs) as knowledge bases (KBs) has shown that language models can act as structured knowledge bases for storing relational facts. However, most existing works only considered the LM-as-KB paradigm in a static setting, which ignores the analysis of temporal dynamics of world knowledge. Furthermore, a basic function of KBs, i.e., the ability to store conflicting information (i.e., 1-N, N-1, and N-M relations), is underexplored. In this paper, we formulate two practical requirements for treating LMs as temporal KBs: (i) The capacity to store temporally-scoped knowledge that contains conflicting information and (ii) the ability to use stored knowledge for temporally-scoped knowledge queries. We introduce a new dataset called LAMA-TK which is aimed at probing temporally-scoped knowledge, and investigate the two above requirements to explore the LM-as-KB paradigm in the temporal domain. On the one hand, experiments show that LMs can memorize millions of temporally-scoped facts with relatively high accuracy and transfer stored knowledge to temporal knowledge queries, thereby expanding the LM-as-KB paradigm to the temporal domain. On the other hand, we show that memorizing conflicting information, which has been neglected by previous works, is still challenging for LMs and hinders the memorization of other unrelated one-to-one relationships.
Are Large Pre-Trained Language Models Leaking Your Personal Information? In this paper, we analyze whether Pre-Trained Language Models (PLMs) are prone to leaking personal information. Specifically, we query PLMs for email addresses with contexts of the email address or prompts containing the owner’s name. We find that PLMs do leak personal information due to memorization. However, since the models are weak at association, the risk of specific personal information being extracted by attackers is low. We hope this work could help the community to better understand the privacy risk of PLMs and bring new insights to make PLMs safe.
In this paper, we propose a self-distillation framework with meta learning (MetaSD) for knowledge graph completion with dynamic pruning, which aims to learn compressed graph embeddings and tackle the long-tail samples. Specifically, we first propose a dynamic pruning technique to obtain a small pruned model from a large source model, where the pruning mask of the pruned model could be updated adaptively per epoch after the model weights are updated. The pruned model is supposed to be more sensitive to difficult-to-memorize samples (e.g., long-tail samples) than the source model. Then, we propose a one-step meta self-distillation method for distilling comprehensive knowledge from the source model to the pruned model, where the two models co-evolve in a dynamic manner during training. In particular, we exploit the performance of the pruned model, which is trained alongside the source model in one iteration, to improve the source model’s knowledge transfer ability for the next iteration via meta learning. Extensive experiments show that MetaSD achieves competitive performance compared to strong baselines, while being 10x smaller than baselines.
Context-dependent text-to-SQL is the task of translating multi-turn questions into database-related SQL queries. Existing methods typically focus on making full use of history context or previously predicted SQL for currently SQL parsing, while neglecting to explicitly comprehend the schema and conversational dependency, such as co-reference, ellipsis and user focus change. In this paper, we propose CQR-SQL, which uses auxiliary Conversational Question Reformulation (CQR) learning to explicitly exploit schema and decouple contextual dependency for multi-turn SQL parsing. Specifically, we first present a schema enhanced recursive CQR method to produce domain-relevant self-contained questions. Secondly, we train CQR-SQL models to map the semantics of multi-turn questions and auxiliary self-contained questions into the same latent space through schema grounding consistency task and tree-structured SQL parsing consistency task, which enhances the abilities of SQL parsing by adequately contextual understanding. At the time of writing, our CQR-SQL achieves new state-of-the-art results on two context-dependent text-to-SQL benchmarks SParC and CoSQL.
Given the recent proliferation of false claims online, there has been a lot of manual fact-checking effort. As this is very time-consuming, human fact-checkers can benefit from tools that can support them and make them more efficient. Here, we focus on building a system that could provide such support. Given an input document, it aims to detect all sentences that contain a claim that can be verified by some previously fact-checked claims (from a given database). The output is a re-ranked list of the document sentences, so that those that can be verified are ranked as high as possible, together with corresponding evidence. Unlike previous work, which has looked into claim retrieval, here we take a document-level perspective. We create a new manually annotated dataset for the task, and we propose suitable evaluation measures. We further experiment with a learning-to-rank approach, achieving sizable performance gains over several strong baselines. Our analysis demonstrates the importance of modeling text similarity and stance, while also taking into account the veracity of the retrieved previously fact-checked claims. We believe that this research would be of interest to fact-checkers, journalists, media, and regulatory authorities.
News articles both shape and reflect public opinion across the political spectrum. Analyzing them for social bias can thus provide valuable insights, such as prevailing stereotypes in society and the media, which are often adopted by NLP models trained on respective data. Recent work has relied on word embedding bias measures, such as WEAT. However, several representation issues of embeddings can harm the measures’ accuracy, including low-resource settings and token frequency differences. In this work, we study what kind of embedding algorithm serves best to accurately measure types of social bias known to exist in US online news articles. To cover the whole spectrum of political bias in the US, we collect 500k articles and review psychology literature with respect to expected social bias. We then quantify social bias using WEAT along with embedding algorithms that account for the aforementioned issues. We compare how models trained with the algorithms on news articles represent the expected social bias. Our results suggest that the standard way to quantify bias does not align well with knowledge from psychology. While the proposed algorithms reduce the gap, they still do not fully match the literature.
This paper examines the encoding of analogy in large-scale pretrained language models, such as BERT and GPT-2. Existing analogy datasets typically focus on a limited set of analogical relations, with a high similarity of the two domains between which the analogy holds. As a more realistic setup, we introduce the Scientific and Creative Analogy dataset (SCAN), a novel analogy dataset containing systematic mappings of multiple attributes and relational structures across dissimilar domains. Using this dataset, we test the analogical reasoning capabilities of several widely-used pretrained language models (LMs). We find that state-of-the-art LMs achieve low performance on these complex analogy tasks, highlighting the challenges still posed by analogy understanding.
Scaling multilingual representation learning beyond the hundred most frequent languages is challenging, in particular to cover the long tail of low-resource languages. We move away from the popular one-for-all multilingual models and focus on training multiple language (family) specific representations, but most prominently enable all languages to still be encoded in the same representational space. We focus on teacher-student training, allowing all encoders to be mutually compatible for bitext mining, and enabling fast learning of new languages. We also combine supervised and self-supervised training, allowing encoders to take advantage of monolingual training data.Our approach significantly outperforms the original LASER encoder. We study very low-resource languages and handle 44 African languages, many of which are not covered by any other model. For these languages, we train sentence encoders and mine bitexts. Adding these mined bitexts yielded an improvement of 3.8 BLEU for NMT into English.
Text-to-SQL parsing tackles the problem of mapping natural language questions to executable SQL queries. In practice, text-to-SQL parsers often encounter various challenging scenarios, requiring them to be generalizable and robust. While most existing work addresses a particular generalization or robustness challenge, we aim to study it in a more comprehensive manner. In specific, we believe that text-to-SQL parsers should be (1) generalizable at three levels of generalization, namely i.i.d., zero-shot, and compositional, and (2) robust against input perturbations. To enhance these capabilities of the parser, we propose a novel TKK framework consisting of Task decomposition, Knowledge acquisition, and Knowledge composition to learn text-to-SQL parsing in stages. By dividing the learning process into multiple stages, our framework improves the parser’s ability to acquire general SQL knowledge instead of capturing spurious patterns, making it more generalizable and robust. Experimental results under various generalization and robustness settings show that our framework is effective in all scenarios and achieves state-of-the-art performance on the Spider, SParC, and CoSQL datasets.
We present EdiT5 - a novel semi-autoregressive text-editing approach designed to combine the strengths of non-autoregressive text-editing and autoregressive decoding. EdiT5 is faster at inference times than conventional sequence-to-sequence (seq2seq) models, while being capable of modeling flexible input-output transformations.This is achieved by decomposing the generation process into three sub-tasks: (1) tagging to decide on the subset of input tokens to be preserved in the output, (2) re-ordering to define their order in the output text, and (3) insertion to infill the missing tokens that are not present in the input. The tagging and re-ordering steps, which are responsible for generating the largest portion of the output, are non-autoregressive, while the insertion uses an autoregressive decoder.Depending on the task, EdiT5 requires significantly fewer autoregressive steps demonstrating speedups of up to 25x when compared to classic seq2seq models. Quality-wise, EdiT5 is initialized with a pre-trained T5 checkpoint yielding comparable performance to T5 in high-resource settings and clearly outperforms it on low-resource settings when evaluated on three NLG tasks: Sentence Fusion, Grammatical Error Correction, and Decontextualization.
We review the state of research on empathy in natural language processing and identify the following issues: (1) empathy definitions are absent or abstract, which (2) leads to low construct validity and reproducibility. Moreover, (3) emotional empathy is overemphasized, skewing our focus to a narrow subset of simplified tasks. We believe these issues hinder research progress and argue that current directions will benefit from a clear conceptualization that includes operationalizing cognitive empathy components. Our main objectives are to provide insight and guidance on empathy conceptualization for NLP research objectives and to encourage researchers to pursue the overlooked opportunities in this area, highly relevant, e.g., for clinical and educational sectors.
Deep neural networks, empowered by pre-trained language models, have achieved remarkable results in natural language understanding (NLU) tasks. However, their performances can drastically deteriorate when logical reasoning is needed. This is because NLU in principle depends on not only analogical reasoning, which deep neural networks are good at, but also logical reasoning. According to the dual-process theory, analogical reasoning and logical reasoning are respectively carried out by System 1 and System 2 in the human brain. Inspired by the theory, we present a novel framework for NLU called Neural-Symbolic Processor (NSP), which performs analogical reasoning based on neural processing and logical reasoning based on both neural and symbolic processing. As a case study, we conduct experiments on two NLU tasks, question answering (QA) and natural language inference (NLI), when numerical reasoning (a type of logical reasoning) is necessary. The experimental results show that our method significantly outperforms state-of-the-art methods in both tasks.
Session-based Social Recommendation (SSR) aims to use users’ social networks and historical sessions to provide more personalized recommendations for the current session.Unfortunately, existing SSR methods have two limitations.First, they do not screen users’ useless social relationships and noisy irrelevant interactions.However, user preferences are mainly affected by several close friends and key interactions.Second, when modeling the current session, they do not take full advantage of user preference information.To tackle these issues, we propose a novel Social-aware Sparse Attention Network for SSR, abbreviated as SSAN.It mainly consists of the Heterogeneous Graph Embedding (HGE) module and the Social-aware Encoder-decoder Network (SEN) module.In the HGE module, we adopt a modified heterogeneous graph neural network, which focuses more on close friends and key historical interactions, to enhance user/item representations. In the SEN module, we use the user representation as a bridge between the Encoder and Decoder to incorporate user preferences when modeling the current session.Extensive experiments on two benchmark datasets demonstrate the superiority of SSAN over the state-of-the-art models.
Adapter Tuning, which freezes the pretrained language models (PLMs) and only fine-tunes a few extra modules, becomes an appealing efficient alternative to the full model fine-tuning. Although computationally efficient, the recent Adapters often increase parameters (e.g. bottleneck dimension) for matching the performance of full model fine-tuning, which we argue goes against their original intention. In this work, we re-examine the parameter-efficiency of Adapter through the lens of network pruning (we name such plug-in concept as SparseAdapter) and find that SparseAdapter can achieve comparable or better performance than standard Adapters when the sparse ratio reaches up to 80%. Based on our findings, we introduce an easy but effective setting “Large-Sparse” to improve the model capacity of Adapters under the same parameter budget. Experiments on five competitive Adapters upon three advanced PLMs show that with proper sparse method (e.g. SNIP) and ratio (e.g. 40%) SparseAdapter can consistently outperform their corresponding counterpart. Encouragingly, with the Large-Sparse setting, we can obtain further appealing gains, even outperforming the full fine-tuning by a large margin.
Quantitative data is important in many domains. Information extraction methods draw structured data from documents. However, the extraction of quantities and their contexts has received little attention in the history of information extraction. In this review, an overview of prior work on measurement extraction is presented. We describe different approaches to measurement extraction and outline the challenges posed by this task. The review concludes with an outline of potential future research. Research strains in measurement extraction tend to be isolated and lack a common terminology. Improvements in numerical reasoning, more extensive datasets, and the consideration of wider contexts may lead to significant improvements in measurement extraction.
Procedural text is a widely used genre that contains many steps of instructions of how to cook a dish or how to conduct a chemical experiment and analyze the procedural text has become a popular task in the NLP field. Since the procedural text can be very long and contains many details, summarizing the whole procedural text or giving an overview for each complicated procedure step can save time for readers and help them to capture the core information in the text. In this paper, we propose the procedural text summarization task with two summarization granularity: step-view and global-view, which summarizes each step in the procedural text separately or gives an overall summary for all steps respectively. To tackle this task, we propose an Entity-State Graph-based Summarizer (ESGS) which is based on state-of-the-art entity state tracking methods and constructs a heterogeneous graph to aggregate contextual information for each procedure. In order to help the summarization model focus on the salient entities, we propose to use the contextualized procedure graph representation to predict the salient entities. Experiments conducted on two datasets verify the effectiveness of our proposed model. Our code and datasets will be released on https://github.com/gsh199449/procedural-summ.
Discriminative pre-trained language models, such as ELECTRA, have achieved promising performances in a variety of general tasks. However, these generic pre-trained models struggle to capture domain-specific knowledge of domain-related tasks. In this work, we propose a novel domain-adaptation method for ELECTRA, which can dynamically select domain-specific tokens and guide the discriminator to emphasize them, without introducing new training parameters. We show that by re-weighting the losses of domain-specific tokens, ELECTRA can be effectively adapted to different domains. The experimental results in both computer science and biomedical domains show that the proposed method can achieve state-of-the-art results on the domain-related tasks.
Automatic Peer-review Aspect Score Prediction (PASP) of academic papers can be a helpful assistant tool for both reviewers and authors. Most existing works on PASP utilize supervised learning techniques. However, the limited number of peer-review data deteriorates the performance of PASP. This paper presents a novel semi-supervised learning (SSL) method that incorporates the Transformer fine-tuning into the Γ-model, a variant of the Ladder network, to leverage contextual features from unlabeled data. Backpropagation simultaneously minimizes the sum of supervised and unsupervised cost functions, avoiding the need for layer-wise pre-training. The experimental results show that our model outperforms the supervised and naive semi-supervised learning baselines. Our source codes are available online.
The successful completion of the hate speech detection task hinges upon the availability of rich and variable labeled data, which is hard to obtain. In this work, we present a new approach for data augmentation that uses as input real unlabelled data, which is carefully selected from online platforms where invited hate speech is abundant. We show that by harvesting and processing this data (in an automatic manner), one can augment existing manually-labeled datasets to improve the classification performance of hate speech classification models. We observed an improvement in F1-score ranging from 2.7% and up to 9.5%, depending on the task (in- or cross-domain) and the model used.
Simultaneous machine translation (SiMT) outputs the translation while receiving the source inputs, and hence needs to balance the received source information and translated target information to make a reasonable decision between waiting for inputs or outputting translation. Previous methods always balance source and target information at the token level, either directly waiting for a fixed number of tokens or adjusting the waiting based on the current token. In this paper, we propose a Wait-info Policy to balance source and target at the information level. We first quantify the amount of information contained in each token, named info. Then during simultaneous translation, the decision of waiting or outputting is made based on the comparison results between the total info of previous target outputs and received source inputs. Experiments show that our method outperforms strong baselines under and achieves better balance via the proposed info.
Simultaneous machine translation (SiMT) starts its translation before reading the whole source sentence and employs either fixed or adaptive policy to generate the target sentence. Compared to the fixed policy, the adaptive policy achieves better latency-quality tradeoffs by adopting a flexible translation policy. If the policy can evaluate rationality before taking action, the probability of incorrect actions will also decrease. However, previous methods lack evaluation of actions before taking them. In this paper, we propose a method of performing the adaptive policy via integrating post-evaluation into the fixed policy. Specifically, whenever a candidate token is generated, our model will evaluate the rationality of the next action by measuring the change in the source content. Our model will then take different actions based on the evaluation results. Experiments on three translation tasks show that our method can exceed strong baselines under all latency.
Sparsity of formal knowledge and roughness of non-ontological construction make sparsity problem particularly prominent in Open Knowledge Graphs (OpenKGs). Due to sparse links, learning effective representation for few-shot entities becomes difficult. We hypothesize that by introducing negative samples, a contrastive learning (CL) formulation could be beneficial in such scenarios. However, existing CL methods model KG triplets as binary objects of entities ignoring the relation-guided ternary propagation patterns and they are too generic, i.e., they ignore zero-shot, few-shot and synonymity problems that appear in OpenKGs. To address this, we propose TernaryCL, a CL framework based on ternary propagation patterns among head, relation and tail. TernaryCL designs Contrastive Entity and Contrastive Relation to mine ternary discriminative features with both negative entities and relations, introduces Contrastive Self to help zero- and few-shot entities learn discriminative features, Contrastive Synonym to model synonymous entities, and Contrastive Fusion to aggregate graph features from multiple paths. Extensive experiments on benchmarks demonstrate the superiority of TernaryCL over state-of-the-art models.
Automatically fixing software bugs is a challenging task. While recent work showed that natural language context is useful in guiding bug-fixing models, the approach required prompting developers to provide this context, which was simulated through commit messages written after the bug-fixing code changes were made. We instead propose using bug report discussions, which are available before the task is performed and are also naturally occurring, avoiding the need for any additional information from developers. For this, we augment standard bug-fixing datasets with bug report discussions. Using these newly compiled datasets, we demonstrate that various forms of natural language context derived from such discussions can aid bug-fixing, even leading to improved performance over using commit messages corresponding to the oracle bug-fixing commits.
Recent studies have shown the impressive efficacy of counterfactually augmented data (CAD) for reducing NLU models’ reliance on spurious features and improving their generalizability. However, current methods still heavily rely on human efforts or task-specific designs to generate counterfactuals, thereby impeding CAD’s applicability to a broad range of NLU tasks. In this paper, we present AutoCAD, a fully automatic and task-agnostic CAD generation framework. AutoCAD first leverages a classifier to unsupervisedly identify rationales as spans to be intervened, which disentangles spurious and causal features. Then, AutoCAD performs controllable generation enhanced by unlikelihood training to produce diverse counterfactuals. Extensive evaluations on multiple out-of-domain and challenge benchmarks demonstrate that AutoCAD consistently and significantly boosts the out-of-distribution performance of powerful pre-trained models across different NLU tasks, which is comparable or even better than previous state-of-the-art human-in-the-loop or task-specific CAD methods.
Classical Chinese poetry has a long history and is a precious cultural heritage of humankind. Displaying the classical Chinese poetry in a visual way, helps to cross cultural barriers in different countries, making it enjoyable for all the people. In this paper, we construct a multi-modal knowledge graph for classical Chinese poetry (PKG), in which the visual information of words in the poetry are incorporated. Then a multi-modal pre-training language model, PKG-Bert, is proposed to obtain the poetry representation with visual information, which bridges the semantic gap between different modalities. PKG-Bert achieves the state-of-the-art performance on the poetry-image retrieval task, showing the effectiveness of incorporating the multi-modal knowledge. The large-scale multi-modal knowledge graph of classical Chinese poetry will be released to promote the researches in classical Chinese culture area.
Recent work on non-autoregressive neural machine translation (NAT) that leverages alignment information to explicitly reduce the modality of target distribution has reported comparable performance with counterparts that tackle multi-modality problem by implicitly modeling dependencies. Effectiveness in handling alignment is vital for models that follow this approach, where a token reordering mechanism is typically involved and plays a vital role. We review the reordering capability of the respective mechanisms in recent NAT models, and our experimental results show that their performance is sub-optimal. We propose to learn a non-autoregressive language model (NALM) based on transformer which can be combined with Viterbi decoding to achieve better reordering performance. We evaluate the proposed NALM using the PTB dataset where sentences with words permuted in different ways are expected to have their ordering recovered. Our empirical results show that the proposed method can outperform the state-of-the-art reordering mechanisms under different word permutation settings, with a 2-27 BLEU improvement, suggesting high potential for word alignment in NAT.
Recent works have revealed that Transformers are implicitly learning the syntactic information in its lower layers from data, albeit is highly dependent on the quality and scale of the training data. However, learning syntactic information from data is not necessary if we can leverage an external syntactic parser, which provides better parsing quality with well-defined syntactic structures. This could potentially improve Transformer’s performance and sample efficiency. In this work, we propose a syntax-guided localized self-attention for Transformer that allows directly incorporating grammar structures from an external constituency parser. It prohibits the attention mechanism to overweight the grammatically distant tokens over close ones. Experimental results show that our model could consistently improve translation performance on a variety of machine translation datasets, ranging from small to large dataset sizes, and with different source languages.
Developing models that can automatically generate detailed code explanation can greatly benefit software maintenance and programming education. However, existing code-to-text generation models often produce only high-level summaries of code that do not capture implementation-level choices essential for these scenarios. To fill in this gap, we propose the code explanation generation task. We first conducted a human study to identify the criteria for high-quality explanatory docstring for code. Based on that, we collected and refined a large-scale code docstring corpus and formulated automatic evaluation metrics that best match human assessments. Finally, we present a multi-stage fine-tuning strategy and baseline models for the task. Our experiments show that (1) our refined training dataset lets models achieve better performance in the explanation generation tasks compared to larger-scale unrefined data (15x larger), and (2) fine-tuned models can generate well-structured long docstrings comparable to human-written ones. We envision our training dataset, human-evaluation protocol, recommended metrics, and fine-tuning strategy can boost future code explanation research. The code and annotated data are available at https://github.com/subercui/CodeExp.
Semantic parsing is an important task that allows to democratize human-computer interaction. One of the most popular text-to-SQL datasets with complex and diverse natural language (NL) questions and SQL queries is Spider. We construct and complement a Spider dataset for Russian, thus creating the first publicly available text-to-SQL dataset for this language. While examining its components - NL questions, SQL queries and databases content - we identify limitations of the existing database structure, fill out missing values for tables and add new requests for underrepresented categories. We select thirty functional test sets with different features that can be used for the evaluation of neural models’ abilities. To conduct the experiments, we adapt baseline architectures RAT-SQL and BRIDGE and provide in-depth query component analysis. On the target language, both models demonstrate strong results with monolingual training and improved accuracy in multilingual scenario. In this paper, we also study trade-offs between machine-translated and manually-created NL queries. At present, Russian text-to-SQL is lacking in datasets as well as trained models, and we view this work as an important step towards filling this gap.
Human reading comprehension often requires reasoning of event semantic relations in narratives, represented by Event-centric Question-Answering (QA). To address event-centric QA, we propose a novel QA model with contrastive learning and invertible event transformation, call TranCLR. Our proposed model utilizes an invertible transformation matrix to project semantic vectors of events into a common event embedding space, trained with contrastive learning, and thus naturally inject event semantic knowledge into mainstream QA pipelines. The transformation matrix is fine-tuned with the annotated event relation types between events that occurred in questions and those in answers, using event-aware question vectors. Experimental results on the Event Semantic Relation Reasoning (ESTER) dataset show significant improvements in both generative and extractive settings compared to the existing strong baselines, achieving over 8.4% gain in the token-level F1 score and 3.0% gain in Exact Match (EM) score under the multi-answer setting. Qualitative analysis reveals the high quality of the generated answers by TranCLR, demonstrating the feasibility of injecting event knowledge into QA model learning. Our code and models can be found at https://github.com/LuJunru/TranCLR.
Multi-Label Few-Shot Aspect Category Detection (FS-ACD) is a new sub-task of aspect-based sentiment analysis, which aims to detect aspect categories accurately with limited training instances. Recently, dominant works use the prototypical network to accomplish this task, and employ the attention mechanism to extract keywords of aspect category from the sentences to produce the prototype for each aspect. However, they still suffer from serious noise problems: (1) due to lack of sufficient supervised data, the previous methods easily catch noisy words irrelevant to the current aspect category, which largely affects the quality of the generated prototype; (2) the semantically-close aspect categories usually generate similar prototypes, which are mutually noisy and confuse the classifier seriously. In this paper, we resort to the label information of each aspect to tackle the above problems, along with proposing a novel Label-Driven Denoising Framework (LDF). Extensive experimental results show that our framework achieves better performance than other state-of-the-art methods.
Visual Entity Linking (VEL) is a task to link regions of images with their corresponding entities in Knowledge Bases (KBs), which is beneficial for many computer vision tasks such as image retrieval, image caption, and visual question answering. While existing tasks in VEL either rely on textual data to complement a multi-modal linking or only link objects with general entities, which fails to perform named entity linking on large amounts of image data. In this paper, we consider a purely Visual-based Named Entity Linking (VNEL) task, where the input only consists of an image. The task is to identify objects of interest (i.e., visual entity mentions) in images and link them to corresponding named entities in KBs. Since each entity often contains rich visual and textual information in KBs, we thus propose three different sub-tasks, i.e., visual to visual entity linking (V2VEL), visual to textual entity linking (V2TEL), and visual to visual-textual entity linking (V2VTEL). In addition, we present a high-quality human-annotated visual person linking dataset, named WIKIPerson. Based on WIKIPerson, we establish a series of baseline algorithms for the solution of each sub-task, and conduct experiments to verify the quality of the proposed datasets and the effectiveness of baseline methods. We envision this work to be helpful for soliciting more works regarding VNEL in the future. The codes and datasets are publicly available at https: //github.com/ict-bigdatalab/VNEL.
Large-scale pretraining is fast becoming the norm in Vision-Language (VL) modeling. However, prevailing VL approaches are limited by the requirement for labeled data and the use of complex multi-step pretraining objectives. We present MAGMA - a simple method for augmenting generative language models with additional modalities using adapter-based finetuning. Building on Frozen, we train a series of VL models that autoregressively generate text from arbitrary combinations of visual and textual input. The pretraining is entirely end-to-end using a single language modeling objective, simplifying optimization compared to previous approaches. Importantly, the language model weights remain unchanged during training, allowing for transfer of encyclopedic knowledge and in-context learning abilities from language pretraining. MAGMA outperforms Frozen on open-ended generative tasks, achieving state of the art results on the OKVQA benchmark and competitive results on a range of other popular VL benchmarks, while pretraining on 0.2 % of the number of samples used to train SimVLM.
Language models (LMs) have been shown to memorize a great deal of factual knowledge contained in their training data. But when an LM generates an assertion, it is often difficult to determine where it learned this information and whether it is true. In this paper, we propose the problem of fact tracing: identifying which training examples taught an LM to generate a particular factual assertion. Prior work on training data attribution (TDA) may offer effective tools for identifying such examples, known as “proponents”. We present the first quantitative benchmark to evaluate this. We compare two popular families of TDA methods — gradient-based and embedding-based — and find that much headroom remains. For example, both methods have lower proponent-retrieval precision than an information retrieval baseline (BM25) that does not have access to the LM at all. We identify key challenges that may be necessary for further improvement such as overcoming the problem of gradient saturation, and also show how several nuanced implementation details of existing neural TDA methods can significantly improve overall fact tracing performance.
Knowledge Graph Question Answering (KGQA) involves retrieving entities as answers from a Knowledge Graph (KG) using natural language queries. The challenge is to learn to reason over question-relevant KG facts that traverse KG entities and lead to the question answers. To facilitate reasoning, the question is decoded into instructions, which are dense question representations used to guide the KG traversals. However, if the derived instructions do not exactly match the underlying KG information, they may lead to reasoning under irrelevant context.Our method, termed ReaRev, introduces a new way to KGQA reasoning with respectto both instruction decoding and execution. To improve instruction decoding, we perform reasoning in an adaptive manner, where KG-aware information is used to iteratively update the initial instructions. To improve instruction execution, we emulate breadth-first search (BFS) with graph neural networks (GNNs). The BFS strategy treats the instructions as a set and allows our method to decide on their execution order on the fly. Experimental results on three KGQA benchmarks demonstrate the ReaRev’s effectiveness compared with previous state-of-the-art, especially when the KG is incomplete or when we tackle complex questions. Our code is publicly available at https://github.com/cmavro/ReaRev_KGQA.
The multimedia communications with texts and images are popular on social media. However, limited studies concern how images are structured with texts to form coherent meanings in human cognition. To fill in the gap, we present a novel concept of cross-modality discourse, reflecting how human readers couple image and text understandings. Text descriptions are first derived from images (named as subtitles) in the multimedia contexts. Five labels – entity-level insertion, projection and concretization and scene-level restatement and extension — are further employed to shape the structure of subtitles and texts and present their joint meanings. As a pilot study, we also build the very first dataset containing over 16K multimedia tweets with manually annotated discourse labels. The experimental results show that trendy multimedia encoders based on multi-head attention (with captions) are unable to well understand cross-modality discourse and additionally modeling texts at the output layer helps yield the-state-of-the-art results.
Recent advances in zero-shot and few-shot learning have shown promise for a scope of research and practical purposes. However, this fast-growing area lacks standardized evaluation suites for non-English languages, hindering progress outside the Anglo-centric paradigm. To address this line of research, we propose TAPE (Text Attack and Perturbation Evaluation), a novel benchmark that includes six more complex NLU tasks for Russian, covering multi-hop reasoning, ethical concepts, logic and commonsense knowledge. The TAPE’s design focuses on systematic zero-shot and few-shot NLU evaluation: (i) linguistic-oriented adversarial attacks and perturbations for analyzing robustness, and (ii) subpopulations for nuanced interpretation. The detailed analysis of testing the autoregressive baselines indicates that simple spelling-based perturbations affect the performance the most, while paraphrasing the input has a more negligible effect. At the same time, the results demonstrate a significant gap between the neural and human baselines for most tasks. We publicly release TAPE (https://tape-benchmark.com) to foster research on robust LMs that can generalize to new tasks when little to no supervision is available.
Knowledge graphs typically contain a large number of entities but often cover only a fraction of all relations between them (i.e., incompleteness). Zero-shot link prediction (ZSLP) is a popular way to tackle the problem by automatically identifying unobserved relations between entities. Most recent approaches use textual features of relations (e.g., surface name or textual descriptions) as auxiliary information to improve the encoded representation. These methods lack robustness as they are bound to support only tokens from a fixed vocabulary and unable to model out-of-vocabulary (OOV) words. Subword units such as character n-grams have the capability of generating more expressive representations for OOV words. Hence, in this paper, we propose a Hierarchical N-gram framework for Zero-Shot Link Prediction (HNZSLP) that leverages character n-gram information for ZSLP. Our approach works by first constructing a hierarchical n-gram graph from the surface name of relations. Subsequently, a new Transformer-based network models the hierarchical n-gram graph to learn a relation embedding for ZSLP. Experimental results show that our proposed HNZSLP method achieves state-of-the-art performance on two standard ZSLP datasets.
Transformer language models such as GPT-2 are difficult to quantize because of outliers in the activations leading to a large quantization error. To adapt to the error, one must use quantization-aware training, which entails a fine-tuning process based on the dataset and the training pipeline identical to those for the original model. Pretrained language models, however, often do not grant access to their datasets and training pipelines, forcing us to rely on arbitrary ones for fine-tuning. In that case, it is observed that quantization-aware training overfits the model to the fine-tuning data. To this end introduced is a quantization adapter (Quadapter), a small set of parameters that are learned to make activations quantization-friendly by scaling them channel-wise.For quantization without overfitting, we introduce a quantization adapter (Quadapter), a small set of parameters that are learned to make activations quantization-friendly by scaling them channel-wise. It keeps the model parameters unchanged. By applying our method to the challenging task of quantizing GPT-2, we demonstrate that it effectively prevents the overfitting and improves the quantization performance.
High-resource languages, such as English, have access to a plethora of datasets with various question-answer types resembling real-world reading comprehension. However, there is a severe lack of diverse and comprehensive question-answering datasets in under-resourced languages like Bangla. The ones available are either translated versions of English datasets with a niche answer format or created by human annotations focusing on a specific domain, question type, or answer type. To address these limitations, this paper introduces BanglaRQA, a reading comprehension-based Bangla question-answering dataset with various question-answer types. BanglaRQA consists of 3,000 context passages and 14,889 question-answer pairs created from those passages. The dataset comprises answerable and unanswerable questions covering four unique categories of questions and three types of answers. In addition, this paper also implemented four different Transformer models for question-answering on the proposed dataset. The best-performing model achieved an overall 62.42% EM and 78.11% F1 score. However, detailed analyses showed that the performance varies across question-answer types, leaving room for substantial improvement of the model performance. Furthermore, we demonstrated the effectiveness of BanglaRQA as a training resource by showing strong results on the bn_squad dataset. Therefore, BanglaRQA has the potential to contribute to the advancement of future research by enhancing the capability of language models. The dataset and codes are available at https://github.com/sartajekram419/BanglaRQA
Given that rich information is hidden behind ubiquitous numbers in text, numerical reasoning over text should be an essential skill of AI systems. To derive precise equations to solve numerical reasoning problems, previous work focused on modeling the structures of equations, and has proposed various structured decoders. Though structure modeling proves to be effective, these structured decoders construct a single equation in a pre-defined autoregressive order, potentially placing an unnecessary restriction on how a model should grasp the reasoning process. Intuitively, humans may have numerous pieces of thoughts popping up in no pre-defined order; thoughts are not limited to the problem at hand, and can even be concerned with other related problems. By comparing diverse thoughts and chaining relevant pieces, humans are less prone to errors. In this paper, we take this inspiration and propose CANTOR, a numerical reasoner that models reasoning steps using a directed acyclic graph where we produce diverse reasoning steps simultaneously without pre-defined decoding dependencies, and compare and chain relevant ones to reach a solution. Extensive experiments demonstrated the effectiveness of CANTOR under both fully-supervised and weakly-supervised settings.
A prominent challenge for modern language understanding systems is the ability to answer implicit reasoning questions, where the required reasoning steps for answering the question are not mentioned in the text explicitly. In this work, we investigate why current models struggle with implicit reasoning question answering (QA) tasks, by decoupling inference of reasoning steps from their execution.We define a new task of implicit relation inference and construct a benchmark, IMPLICITRELATIONS, where given a question, a model should output a list of concept-relation pairs, where the relations describe the implicit reasoning steps required for answering the question.Using IMPLICITRELATIONS, we evaluate models from the GPT-3 family and find that, while these models struggle on the implicit reasoning QA task, they often succeed at inferring implicit relations.This suggests that the challenge in implicit reasoning questions does not stem from the need to plan a reasoning strategy alone, but to do it while also retrieving and reasoning over relevant information.
Recent works suggest that transformer models are capable of multi-tasking on diverse NLP tasks and adapt to new tasks efficiently. However, the potential of these multi-task models may be limited as they use the same set of parameters for all tasks. In contrast, humans tackle tasks in a more flexible way, by making proper presumptions on what skills and knowledge are relevant and executing only the necessary computations. Inspired by this, we propose to use task-level mixture-of-expert models, which has a collection of transformer layers (i.e., experts) and a router component to choose among these experts dynamically and flexibly. We find that these models help improve the average performance gain (ARG) metric by 2.6% when adapting to unseen tasks in few-shot settings, and by 5.6% in zero-shot generalization settings. Further, we show that the learned routing decisions and experts partly rediscover human categorization of NLP tasks – certain experts are strongly associated with extractive tasks, some with classification tasks, and some with tasks requiring world knowledge.
We show that the l2 norm of a static sense embedding encodes information related to the frequency of that sense in the training corpus used to learn the sense embeddings. This finding can be seen as an extension of a previously known relationship for word embeddings to sense embeddings. Our experimental results show that in spite of its simplicity, the l2 norm of sense embeddings is a surprisingly effective feature for several word sense related tasks such as (a) most frequent sense prediction, (b) word-in-context (WiC), and (c) word sense disambiguation (WSD). In particular, by simply including the l2 norm of a sense embedding as a feature in a classifier, we show that we can improve WiC and WSD methods that use static sense embeddings.
In the past few years, pre-trained BERT has become one of the most popular deep-learning language models due to their remarkable performance in natural language processing (NLP) tasks. However, the superior performance of BERT comes at the cost of high computational and memory complexity, hindering its envisioned widespread deployment in edge devices with limited computing resources. Binarization can alleviate these limitations by reducing storage requirements and improving computing performance. However, obtaining a comparable accuracy performance for binary BERT w.r.t. its full-precision counterpart is still a difficult task. We observe that direct binarization of pre-trained BERT provides a poor initialization during the fine-tuning phase, making the model incapable of achieving a decent accuracy on downstream tasks. Based on this observation, we put forward the following hypothesis: partially randomly-initialized BERT with binary weights and activations can reach to a decent accuracy performance by distilling knowledge from the its full-precision counterpart. We show that BERT with pre-trained embedding layer and randomly-initialized encoder is a smoking gun for this hypothesis. We identify the smoking gun through a series of experiments and show that it yields a new set of state-of-the-art results on the GLUE and SQuAD benchmarks.
One of the most impressive results of recent NLP history is the ability of pre-trained language models to solve new tasks in a zero-shot setting. To achieve this, NLP tasks are framed as natural language prompts, generating a response indicating the predicted output. Nonetheless, the performance in such settings often lags far behind its supervised counterpart, suggesting a large space for potential improvement. In this paper, we explore methods to utilize unlabeled data to improve zero-shot performance. Specifically, we take advantage of the fact that multiple prompts can be used to specify a single task, and propose to regularize prompt consistency, encouraging consistent predictions over this diverse set of prompts. Our method makes it possible to fine-tune the model either with extra unlabeled training data, or directly on test input at inference time in an unsupervised manner. In experiments, our approach outperforms the state-of-the-art zero-shot learner, T0, on 9 out of 11 datasets across 4 NLP tasks by up to 10.6 absolute points in terms of accuracy. The gains are often attained with a small number of unlabeled examples.
Collecting and annotating task-oriented dialogues is time-consuming and costly. Thus, zero and few shot learning for dialogue tasks presents an exciting opportunity. In this work, we propose an in-context (IC) learning framework for zero-shot and few-shot learning dialogue state tracking (DST), where a large pretrained language model (LM) takes a test instance and a few exemplars as input, and directly decodes the dialogue state without any parameter updates. This approach is more flexible and scalable than prior DST work when adapting to new domains and scenarios. To better leverage a tabular domain description in the LM prompt, we reformulate DST into a text-to-SQL problem. We also propose a novel approach to retrieve annotated dialogues as exemplars. Empirical results on MultiWOZ show that our method IC-DST substantially outperforms previous fine-tuned state-of-the-art models in few-shot settings. In addition, we test IC-DST in zero-shot settings, in which the model only takes a fixed task instruction as input, finding that it outperforms previous zero-shot methods by a large margin.
Combining the visual modality with pretrained language models has been surprisingly effective for simple descriptive tasks such as image captioning. More general text generation however remains elusive. We take a step back and ask: How do these models work for more complex generative tasks, i.e. conditioning on both text and images? Are multimodal models simply visually adapted language models, or do they combine they reason jointly over modalities?We investigate these questions in the context of self-rationalization (jointly generating task labels/answers and free-text explanations) of three tasks: (i) visual question answering in VQA-X, (ii) visual commonsense reasoning in VCR, and (iii) visual-textual entailment in E-SNLI-VE. We show that recent unimodal advances, CLIP image representations and scaling of language models, do not consistently improveself-rationalization in multimodal tasks. We find that no single model type works universally best across tasks, datasets, and finetuning data sizes. Our findings motivate the need for novel general backbones that move text generation from images and text beyond image captioning.
Language use changes over time, and this impacts the effectiveness of NLP systems. This phenomenon is even more prevalent in social media data during crisis events where meaning and frequency of word usage may change over the course of days. Contextual language models fail to adapt temporally, emphasizing the need for temporal adaptation in models which need to be deployed over an extended period of time. While existing approaches consider data spanning large periods of time (from years to decades), shorter time spans are critical for crisis data. We quantify temporal degradation for this scenario and propose methods to cope with performance loss by leveraging techniques from domain adaptation. To the best of our knowledge, this is the first effort to explore effects of rapid language change driven by adversarial adaptations, particularly during natural and human-induced disasters. Through extensive experimentation on diverse crisis datasets, we analyze under what conditions our approaches outperform strong baselines while highlighting the current limitations of temporal adaptation methods in scenarios where access to unlabeled data is scarce.
The field of Deep Learning (DL) has undergone explosive growth during the last decade, with a substantial impact on Natural Language Processing (NLP) as well. Yet, compared to more established disciplines, a lack of common experimental standards remains an open challenge to the field at large. Starting from fundamental scientific principles, we distill ongoing discussions on experimental standards in NLP into a single, widely-applicable methodology. Following these best practices is crucial to strengthen experimental evidence, improve reproducibility and enable scientific progress. These standards are further collected in a public repository to help them transparently adapt to future needs.
Anaphora resolution is an important task for information extraction across a range of languages, text genres, and domains, motivating the need for methods that do not require large annotated datasets. In-context learning has emerged as a promising approach, yet there are a number of challenges in applying in-context learning to resolve anaphora. For example, encoding a single in-context demonstration that consists of: an anaphor, a paragraph-length context, and a list of corresponding antecedents, requires conditioning a language model on a long sequence of tokens, limiting the number of demonstrations per prompt.In this paper, we present Mice (Mixtures of In-Context Experts), which we demonstrate is effective for few-shot anaphora resolution in scientific protocols. Given only a handful of training examples, Mice combines the predictions of hundreds of in-context experts, yielding a 30% increase in F1 score over a competitive prompt retrieval baseline. Furthermore, we show Mice can be used to train compact student models without sacrificing performance. As far as we are aware, this is the first work to present experimental results demonstrating the effectiveness of in-context learning on the task of few-shot anaphora resolution in scientific protocols.
We investigate the problem of determining the predictive confidence (or, conversely, uncertainty) of a neural classifier through the lens of low-resource languages. By training models on sub-sampled datasets in three different languages, we assess the quality of estimates from a wide array of approaches and their dependence on the amount of available data. We find that while approaches based on pre-trained models and ensembles achieve the best results overall, the quality of uncertainty estimates can surprisingly suffer with more data. We also perform a qualitative analysis of uncertainties on sequences, discovering that a model’s total uncertainty seems to be influenced to a large degree by its data uncertainty, not model uncertainty. All model implementations are open-sourced in a software package.
Contrastive representation learning has gained much attention due to its superior performance in learning representations from both image and sequential data. However, the learned representations could potentially lead to performance disparities in downstream tasks, such as increased silencing of underrepresented groups in toxicity comment classification. In light of this challenge, in this work, we study learning fair representations that satisfy a notion of fairness known as equalized odds for text classification via contrastive learning. Specifically, we first theoretically analyze the connections between learning representations with a fairness constraint and conditional supervised contrastive objectives, and then propose to use conditional supervised contrastive objectives to learn fair representations for text classification. We conduct experiments on two text datasets to demonstrate the effectiveness of our approaches in balancing the trade-offs between task performance and bias mitigation among existing baselines for text classification. Furthermore, we also show that the proposed methods are stable in different hyperparameter settings.
Named geographic entities (geo-entities for short) are the building blocks of many geographic datasets. Characterizing geo-entities is integral to various application domains, such as geo-intelligence and map comprehension, while a key challenge is to capture the spatial-varying context of an entity. We hypothesize that we shall know the characteristics of a geo-entity by its surrounding entities, similar to knowing word meanings by their linguistic context. Accordingly, we propose a novel spatial language model, SpaBERT, which provides a general-purpose geo-entity representation based on neighboring entities in geospatial data. SpaBERT extends BERT to capture linearized spatial context, while incorporating a spatial coordinate embedding mechanism to preserve spatial relations of entities in the 2-dimensional space. SpaBERT is pretrained with masked language modeling and masked entity prediction tasks to learn spatial dependencies. We apply SpaBERT to two downstream tasks: geo-entity typing and geo-entity linking. Compared with the existing language models that do not use spatial context, SpaBERT shows significant performance improvement on both tasks. We also analyze the entity representation from SpaBERT in various settings and the effect of spatial coordinate embedding.
In task-oriented dialogue systems, response generation from meaning representations (MRs) often suffers from limited training examples, due to the high cost of annotating MR-to-Text pairs. Previous works on self-training leverage fine-tuned conversational models to automatically generate pseudo-labeled MR-to-Text pairs for further fine-tuning. However, some self-augmented data may be noisy or uninformative for the model to learn from. In this work, we propose a two-phase self-augmentation procedure to generate high-quality pseudo-labeled MR-to-Text pairs: the first phase selects the most informative MRs based on model’s prediction uncertainty; with the selected MRs, the second phase generates accurate responses by aggregating multiple perturbed latent representations from each MR. Empirical experiments on two benchmark datasets, FewShotWOZ and FewShotSGD, show that our method generally outperforms existing self-training methods on both automatic and human evaluations.
While standardization is a well-established activity in other scientific fields such as telecommunications, networks or multimedia, in the field of AI and more specifically NLP it is still at its dawn. In this paper, we explore how various aspects of NLP (evaluation, data, tasks...) lack standards and how that can impact science, but also the society, the industry, and regulations. We argue that the numerous initiatives to rationalize the field and establish good practices are only the first step, and developing formal standards remains needed to bring further clarity to NLP research and industry, at a time where this community faces various crises regarding ethics or reproducibility. We thus encourage NLP researchers to contribute to existing and upcoming standardization projects, so that they can express their needs and concerns, while sharing their expertise.
Next-word predictions from autoregressive neural language models show remarkable sensitivity to syntax. This work evaluates the extent to which this behavior arises as a result of a learned ability to maintain implicit representations of incremental syntactic structures. We extend work in syntactic probing to the incremental setting and present several probes for extracting incomplete syntactic structure (operationalized through parse states from a stack-based parser) from autoregressive language models. We find that our probes can be used to predict model preferences on ambiguous sentence prefixes and causally intervene on model representations and steer model behavior. This suggests implicit incremental syntactic inferences underlie next-word predictions in autoregressive neural language models.
For users to trust model predictions, they need to understand model outputs, particularly their confidence — calibration aims to adjust (calibrate) models’ confidence to match expected accuracy. We argue that the traditional calibration evaluation does not promote effective calibrations: for example, it can encourage always assigning a mediocre confidence score to all predictions, which does not help users distinguish correct predictions from wrong ones. Building on those observations, we propose a new calibration metric, MacroCE, that better captures whether the model assigns low confidence to wrong predictions and high confidence to correct predictions. Focusing on the practical application of open-domain question answering, we examine conventional calibration methods applied on the widely-used retriever-reader pipeline, all of which do not bring significant gains under our new MacroCE metric. Toward better calibration, we propose a new calibration method (ConsCal) that uses not just final model predictions but whether multiple model checkpoints make consistent predictions. Altogether, we provide an alternative view of calibration along with a new metric, re-evaluation of existing calibration methods on our metric, and proposal of a more effective calibration method.
Novel inverted index-based learned sparse ranking models provide more effective, but less efficient, retrieval performance compared to traditional ranking models like BM25. In this paper, we introduce a technique we call postings clipping to improve the query efficiency of learned representations. Our technique amplifies the benefit of dynamic pruning query processing techniques by accounting for changes in term importance distributions of learned ranking models. The new clipping mechanism accelerates top-k retrieval by up to 9.6X without any loss in effectiveness.
Traditional multi-task learning architectures learn a single model across multiple tasks through a shared encoder followed by task-specific decoders. Learning these models often requires specialized training algorithms that address task-conflict in the shared parameter updates, which otherwise can lead to negative transfer. A new type of multi-task learning within NLP homogenizes multi-task architectures as a shared encoder and language model decoder, which does surprisingly well across a range of diverse tasks. Does this new architecture suffer from task-conflicts that require specialized training algorithms? We study how certain factors in the shift towards text-to-text models affects multi-task conflict and negative transfer, finding that both directional conflict and transfer are surprisingly constant across architectures.
Static subword tokenization algorithms have been an essential component of recent works on language modeling. However, their static nature results in important flaws that degrade the models’ downstream performance and robustness. In this work, we propose MANTa, a Module for Adaptive Neural TokenizAtion. MANTa is a differentiable tokenizer trained end-to-end with the language model. The resulting system offers a trade-off between the expressiveness of byte-level models and the speed of models trained using subword tokenization. In addition, our tokenizer is highly explainable since it produces an explicit segmentation of sequences into blocks. We evaluate our pre-trained model on several English datasets from different domains as well as on synthetic noise. We find that MANTa improves robustness to character perturbations and out-of-domain data. We then show that MANTa performs comparably to other models on the general-domain GLUE benchmark. Finally, we show that it is considerably faster than strictly byte-level models.
NLP offers a myriad of opportunities to support mental health research. However, prior work has almost exclusively focused on social media data, for which diagnoses are difficult or impossible to validate. We present a first-of-its-kind dataset of manually transcribed interactions with people clinically diagnosed with bipolar disorder and schizophrenia, as well as healthy controls. Data was collected through validated clinical tasks and paired with diagnostic measures. We extract 100+ temporal, sentiment, psycholinguistic, emotion, and lexical features from the data and establish classification validity using a variety of models to study language differences between diagnostic groups. Our models achieve strong classification performance (maximum F1=0.93-0.96), and lead to the discovery of interesting associations between linguistic features and diagnostic class. It is our hope that this dataset will offer high value to clinical and NLP researchers, with potential for widespread broader impacts.
Multi-hop Question Answering (QA) is a challenging task since it requires an accurate aggregation of information from multiple context paragraphs and a thorough understanding of the underlying reasoning chains. Recent work in multi-hop QA has shown that performance can be boosted by first decomposing the questions into simpler, single-hop questions. In this paper, we explore one additional utility of the multi-hop decomposition from the perspective of explainable NLP: to create explanation by probing a neural QA model with them. We hypothesize that in doing so, users will be better able to predict when the underlying QA system will give the correct answer. Through human participant studies, we verify that exposing the decomposition probes and answers to the probes to users can increase their ability to predict system performance on a question instance basis. We show that decomposition is an effective form of probing QA systems as well as a promising approach to explanation generation. In-depth analyses show the need for improvements in decomposition systems.
An adversarial attack generates harmful text that fools a target model. More dangerously, this text is unrecognizable by humans. Existing work detects adversarial text and corrects a target’s prediction by identifying perturbed words and changing them into their synonyms, but many benign words are also changed. In this paper, we directly detect adversarial text, correct the prediction, and suggest perturbed words by checking the change in the hard labels from the target’s predictions after replacing a word with its transformation using a model that we call CheckHARD. The experiments demonstrate that CheckHARD outperforms existing work on various attacks, models, and datasets.
An increasingly prevalent problem for intelligent technologies is text safety, as uncontrolled systems may generate recommendations to their users that lead to injury or life-threatening consequences. However, the degree of explicitness of a generated statement that can cause physical harm varies. In this paper, we distinguish types of text that can lead to physical harm and establish one particularly underexplored category: covertly unsafe text. Then, we further break down this category with respect to the system’s information and discuss solutions to mitigate the generation of text in each of these subcategories. Ultimately, our work defines the problem of covertly unsafe language that causes physical harm and argues that this subtle yet dangerous issue needs to be prioritized by stakeholders and regulators. We highlight mitigation strategies to inspire future researchers to tackle this challenging problem and help improve safety within smart systems.
Humor plays an important role in our daily life, as it is an essential and fascinating element in the communication between persons. Therefore, how to recognize punchlines from the dialogue, i.e. conversational humor recognition, has attracted much interest of computational linguistics communities. However, most existing work attempted to understand the conversational humor by analyzing the contextual information of the dialogue, but neglected the character of the interlocutor, such as age, gender, occupation, and so on. For instance, the same utterance could bring out humorous from a serious person, but may be a plain expression from a naive person. To this end, this paper proposes a Character Fusion Conversational Humor Recognition model (CFCHR) to explore character information to recognize conversational humor. CFCHR utilizes a multi-task learning framework that unifies two highly pertinent tasks, i.e., character extraction and punchline identification. Based on deep neural networks, we trained both tasks jointly by sharing weight to extract the common and task-invariant features while each task could still learn its task-specific features. Experiments were conducted on Chinese sitcoms corpus, which consisted of 12,677 utterances from 22 characters. The experimental results demonstrated that CFCHR could achieve 33.08% improvements in terms of F1-score over some strong baselines, and proved the effectiveness of the character information to identify the punchlines.
Click behaviors are widely used for learning news recommendation models, but they are heavily affected by the biases brought by the news display positions. It is important to remove position biases to train unbiased recommendation model and capture unbiased user interest. In this paper, we propose a news recommendation method named DebiasGAN that can effectively alleviate position biases via adversarial learning. The core idea is modeling the personalized effect of position bias on click behaviors in a candidate-aware way, and learning debiased candidate-aware user embeddings from which the position information cannot be discriminated. More specifically, we use a bias-aware click model to capture the effect of position bias on click behaviors, and use a bias-invariant click model with random candidate positions to estimate the ideally unbiased click scores. We apply adversarial learning to the embeddings learned by the two models to help the bias-invariant click model capture debiased user interest. Experimental results on two real-world datasets show that DebiasGAN effectively improves news recommendation by eliminating position biases.
Sentence summarization shortens given texts while maintaining core contents of the texts. Unsupervised approaches have been studied to summarize texts without ground-truth summaries. However, recent unsupervised models are extractive, which remove words from texts and thus they are less flexible than abstractive summarization. In this work, we devise an abstractive model based on reinforcement learning without ground-truth summaries. We formulate the unsupervised summarization based on the Markov decision process with rewards representing the summary quality. To further enhance the summary quality, we develop a multi-summary learning mechanism that generates multiple summaries with varying lengths for a given text, while making the summaries mutually enhance each other. Experimental results show that the proposed model substantially outperforms both abstractive and extractive models, yet frequently generating new words not contained in input texts.
Multilingual pretrained language models (mPLMs) have shown their effectiveness in multilingual word alignment induction. However, these methods usually start from mBERT or XLM-R. In this paper, we investigate whether multilingual sentence Transformer LaBSE is a strong multilingual word aligner. This idea is non-trivial as LaBSE is trained to learn language-agnostic sentence-level embeddings, while the alignment extraction task requires the more fine-grained word-level embeddings to be language-agnostic. We demonstrate that the vanilla LaBSE outperforms other mPLMs currently used in the alignment task, and then propose to finetune LaBSE on parallel corpus for further improvement. Experiment results on seven language pairs show that our best aligner outperforms previous state-of-the-art models of all varieties. In addition, our aligner supports different language pairs in a single model, and even achieves new state-of-the-art on zero-shot language pairs that does not appear in the finetuning process.
Counterfactual data augmentation (CDA) – i.e., adding minimally perturbed inputs during training – helps reduce model reliance on spurious correlations and improves generalization to out-of-distribution (OOD) data. Prior work on generating counterfactuals only considered restricted classes of perturbations, limiting their effectiveness. We present Counterfactual Generation via Retrieval and Editing (CORE), a retrieval-augmented generation framework for creating diverse counterfactual perturbations for CDA. For each training example, CORE first performs a dense retrieval over a task-related unlabeled text corpus using a learned bi-encoder and extracts relevant counterfactual excerpts. CORE then incorporates these into prompts to a large language model with few-shot learning capabilities, for counterfactual editing. Conditioning language model edits on naturally occurring data results in more diverse perturbations. Experiments on natural language inference and sentiment analysis benchmarks show that CORE counterfactuals are more effective at improving generalization to OOD data compared to other DA approaches. We also show that the CORE retrieval framework can be used to encourage diversity in manually authored perturbations.
Conversation disentanglement aims to group utterances into detached sessions, which is a fundamental task in processing multi-party conversations. Existing methods have two main drawbacks. First, they overemphasize pairwise utterance relations but pay inadequate attention to the utterance-to-context relation modeling. Second, huge amount of human annotated data is required for training, which is expensive to obtain in practice. To address these issues, we propose a general disentangle model based on bi-level contrastive learning. It brings closer utterances in the same session while encourages each utterance to be near its clustered session prototypes in the representation space. Unlike existing approaches, our disentangle model works in both supervised setting with labeled data and unsupervised setting when no such data is available. The proposed method achieves new state-of-the-art performance on both settings across several public datasets.
Retrieval-enhanced language models (LMs), which condition their predictions on text retrieved from large external datastores, have recently shown significant perplexity improvements compared to standard LMs. One such approach, the kNN-LM, interpolates any existing LM’s predictions with the output of a k-nearest neighbors model and requires no additional training. In this paper, we explore the importance of lexical and semantic matching in the context of items retrieved by kNN-LM. We find two trends: (1) the presence of large overlapping n-grams between the datastore and evaluation set plays an important factor in strong performance, even when the datastore is derived from the training data; and (2) the kNN-LM is most beneficial when retrieved items have high semantic similarity with the query. Based on our analysis, we define a new formulation of the kNN-LM that uses retrieval quality to assign the interpolation coefficient. We empirically measure the effectiveness of our approach on two English language modeling datasets, Wikitext-103 and PG-19. Our re-formulation of the kNN-LM is beneficial in both cases, and leads to nearly 4% improvement in perplexity on the Wikitext-103 test set.
This paper proposes StuBot, a text-based conversational agent that provides adaptive feedback for learning by teaching. StuBot first asks the users to teach the learning content by summarizing and explaining it in their own words. After the users inputted the explanation text for teaching, StuBot uses a machine reading comprehension (MRC) engine to provide adaptive feedback with further questions about the insufficient parts of the explanation text. We conducted a within-subject study to evaluate the effectiveness of adaptive feedback by StuBot. Both the quantitative and qualitative results showed that learning by teaching with adaptive feedback can improve learning performance, immersion, and overall experience.
Contrastive learning has been demonstrated to be effective in enhancing pre-trained language models (PLMs) to derive superior universal sentence embeddings. However, existing contrastive methods still have two limitations. Firstly, previous works may acquire poor performance under domain shift settings, thus hindering the application of sentence representations in practice. We attribute this low performance to the over-parameterization of PLMs with millions of parameters. To alleviate it, we propose PromCSE (Prompt-based Contrastive Learning for Sentence Embeddings), which only trains small-scale Soft Prompt (i.e., a set of trainable vectors) while keeping PLMs fixed. Secondly, the commonly used NT-Xent loss function of contrastive learning does not fully exploit hard negatives in supervised learning settings. To this end, we propose to integrate an Energy-based Hinge loss to enhance the pairwise discriminative power, inspired by the connection between the NT-Xent loss and the Energy-based Learning paradigm. Empirical results on seven standard semantic textual similarity (STS) tasks and a domain-shifted STS task both show the effectiveness of our method compared with the current state-of-the-art sentence embedding models.
Video language pre-training methods have mainly adopted sparse sampling techniques to alleviate the temporal redundancy of videos. Though effective, sparse sampling still suffers inter-modal redundancy: visual redundancy and textual redundancy. Compared with highly generalized text, sparsely sampled frames usually contain text-independent portions, called visual redundancy. Sparse sampling is also likely to miss important frames corresponding to some text portions, resulting in textual redundancy. Inter-modal redundancy leads to a mismatch of video and text information, hindering the model from better learning the shared semantics across modalities. To alleviate it, we propose Redundancy-aware Video-language Pre-training. We design a redundancy measurement of video patches and text tokens by calculating the cross-modal minimum dis-similarity. Then, we penalize the high-redundant video patches and text tokens through a proposed redundancy-aware contrastive learning. We evaluate our method on four benchmark datasets, MSRVTT, MSVD, DiDeMo, and LSMDC, achieving a significant improvement over the previous state-of-the-art results.
It is notoriously difficult to implement end-to-end speech translation (E2E-ST) model because of the task complexity and data scarcity. Existing techniques often attempt to carry out implicit knowledge transfer from machine translation (MT) to ST model by imposing various constraints. However, in this transfer scenario, a significant problem is that the performance of the MT will drop significantly and the final transfer effect is also restricted. In this article, we recommend Fine and Coarse Granularity Contrastive Learning (FCGCL), which conduct explicit knowledge transfer from MT to ST model. Specially, we ensure through multi granularity contrastive learning that inputs with similar semantic between different modalities are encoded closely in the shared semantic space while inputs with different semantics are kept apart. Experiments on the MuST-C datasets on all 8 languages and further analysis show that our method can effectively improve the E2E-ST performance and achieves an average BLEU of 29.0.
Contrastive learning has been extensively studied in sentence embedding learning, which assumes that the embeddings of different views of the same sentence are closer. The constraint brought by this assumption is weak, and a good sentence representation should also be able to reconstruct the original sentence fragments. Therefore, this paper proposes an information-aggregated contrastive learning framework for learning unsupervised sentence embeddings, termed InfoCSE.InfoCSE forces the representation of [CLS] positions to aggregate denser sentence information by introducing an additional Masked language model task and a well-designed network. We evaluate the proposed InfoCSE on several benchmark datasets w.r.t the semantic text similarity (STS) task. Experimental results show that InfoCSE outperforms SimCSE by an average Spearman correlation of 2.60% on BERT-base, and 1.77% on BERT-large, achieving state-of-the-art results among unsupervised sentence representation learning methods.
Pre-trained language models have demonstrated impressive performance in both natural language processing and program understanding, which represent the input as a token sequence without explicitly modeling its structure. Some prior works show that pre-trained language models can capture the syntactic rules of natural languages without finetuning on syntax understanding tasks. However, there is limited understanding of how well pre-trained models understand the code structure so far. In this work, we perform the first thorough benchmarking of the state-of-the-art pre-trained models for identifying the syntactic structures of programs. Specifically, we introduce CodeSyntax, a large-scale dataset of programs annotated with the syntactic relationships in their corresponding abstract syntax trees. Our key observation is that pre-training on massive code data does not result in decent code syntax understanding. In fact, these pre-trained programming language models fail to match the performance of naive baselines based on positional offsets and keywords. We also present a natural language benchmark to highlight the differences between natural languages and programming languages in terms of understanding corresponding syntactic structures. Our findings point out key limitations of existing pre-training methods and suggest the importance of modeling syntactic structures for the programming language.
This work extends the current quotation recommendation task to a more realistic quotation recommender system that learns to predict when to quote and what to quote jointly. The system consists of three modules (tasks), a prediction module to predict whether to quote given conversation contexts, a recommendation module to recommend suitable quotations and a generation module generating quotations or sentences in ordinary language to continue the conversation. We benchmark several competitive models for the two newly introduced tasks (i.e., when-to-quote and what-to-continue). For quotation recommendation, compared with previous work that is either generation-based or ranking-based recommendation, we propose a novel framework with mutual promotion of generation module and ranking-based recommendation module. Experiments show that our framework achieves significantly better performance than baselines on two datasets. Further experiments and analyses validate the effectiveness of the proposed mechanisms and get a better understanding of the quotation recommendation task.
Commonsense knowledge has been widely considered for building intelligent open-domain dialogue agents, aiming to generate meaningful and diverse responses. Previous works in this field usually lack the ability to effectively obtain and utilize auxiliary commonsense from the external visual world. In this paper, we argue that exploiting logical information in images related to context can be effective to enrich and steer the generation process. In view of this, we propose VICTOR, a context-relevant VIsual Commonsense enhanced dialogue generaTOR for generating coherent and informative responses. To obtain the associated visual commonsense, we devise a novel approach that expands topic words on the knowledge graph and maps them into daily scenarios. During the generation, the model adopts multimodal fusion mechanism to integrate visual and textual information, and adaptively combine their decoding distributions for better response generation. The experimental results on two public datasets show that our proposed method outperforms the latest competitive methods in terms of coherence and diversity.
Different methods have been proposed to develop meta-embeddings from a given set of source embeddings. However, the source embeddings can contain unfair gender-related biases, and how these influence the meta-embeddings has not been studied yet.We study the gender bias in meta-embeddings created under three different settings:(1) meta-embedding multiple sources without performing any debiasing (Multi-Source No-Debiasing),(2) meta-embedding multiple sources debiased by a single method (Multi-Source Single-Debiasing), and(3) meta-embedding a single source debiased by different methods (Single-Source Multi-Debiasing).Our experimental results show that meta-embedding amplifies the gender biases compared to input source embeddings.We find that debiasing not only the sources but also their meta-embedding is needed to mitigate those biases.Moreover, we propose a novel debiasing method based on meta-embedding learning where we use multiple debiasing methods on a single source embedding and then create a single unbiased meta-embedding.
Word alignment is to find translationally equivalent words between source and target sentences. Previous work has demonstrated that self-training can achieve competitive word alignment results. In this paper, we propose to use word alignments generated by a third-party word aligner to supervise the neural word alignment training. Specifically, source word and target word of each word pair aligned by the third-party aligner are trained to be close neighbors to each other in the contextualized embedding space when fine-tuning a pre-trained cross-lingual language model. Experiments on the benchmarks of various language pairs show that our approach can surprisingly do self-correction over the third-party supervision by finding more accurate word alignments and deleting wrong word alignments, leading to better performance than various third-party word aligners, including the currently best one. When we integrate all supervisions from various third-party aligners, we achieve state-of-the-art word alignment performances, with averagely more than two points lower alignment error rates than the best third-party aligner.We released our code at https://github.com/sdongchuanqi/Third-Party-Supervised-Aligner.
Fact verification is an essential tool to mitigate the spread of false information online, which has gained a widespread attention recently. However, a fact verification in the question-answering dialogue is still underexplored. In this paper, we propose a neural network based approach called question-answering dialogue based fact verification with mixture of experts (QaDialMoE). It exploits questions and evidence effectively in the verification process and can significantly improve the performance of fact verification. Specifically, we exploit the mixture of experts to focus on various interactions among responses, questions and evidence. A manager with an attention guidance module is implemented to guide the training of experts and assign a reasonable attention score to each expert. A prompt module is developed to generate synthetic questions that make our approach more generalizable. Finally, we evaluate the QaDialMoE and conduct a comparative study on three benchmark datasets. The experimental results demonstrate that our QaDialMoE outperforms previous approaches by a large margin and achieves new state-of-the-art results on all benchmarks. This includes the accuracy improvements on the HEALTHVER as 84.26%, the FAVIQ A dev set as 78.7%, the FAVIQ R dev set as 86.1%, test set as 86.0%, and the COLLOQUIAL as 89.5%. To our best knowledge, this is the first work to investigate a question-answering dialogue based fact verification, and achieves new state-of-the-art results on various benchmark datasets.
With the popularity of online social media, massive-scale multimodal information has brought new challenges to traditional Named Entity Disambiguation (NED) tasks. Recently, Multimodal Named Entity Disambiguation (MNED) has been proposed to link ambiguous mentions with the textual and visual contexts to a predefined knowledge graph. Existing attempts usually perform MNED by annotating multimodal mentions and adding multimodal features to traditional NED models. However, these studies may suffer from 1) failing to model multimodal information at the knowledge level, and 2) lacking multimodal annotation data against the large-scale unlabeled corpus. In this paper, we explore a pioneer study on leveraging multimodal knowledge learning to address the MNED task. Specifically, we first harvest multimodal knowledge in the Meta-Learning way, which is much easier than collecting ambiguous mention corpus. Then we design a knowledge-guided transfer learning strategy to extract unified representation from different modalities. Finally, we propose an Interactive Multimodal Learning Network (IMN) to fully utilize the multimodal information on both the mention and knowledge sides. Extensive experiments conducted on two public MNED datasets demonstrate that the proposed method achieves improvements over the state-of-the-art multimodal methods.
Using prompts to explore the knowledge contained within pre-trained language models for downstream tasks has now become an active topic. Current prompt tuning methods mostly convert the downstream tasks to masked language modeling problems by adding cloze-style phrases and mapping all labels to verbalizations with fixed length, which has proven effective for tasks with simple label spaces. However, when applied to relation classification exhibiting complex label spaces, vanilla prompt tuning methods may struggle with label verbalizations with arbitrary lengths due to rigid prompt restrictions. Inspired by the text infilling task for pre-training generative models that can flexibly predict missing spans, we propose a novel generative prompt tuning method to reformulate relation classification as an infilling problem, which frees our approach from limitations of current prompt based approaches and thus fully exploits rich semantics of entity and relation types. In addition, we design entity-guided decoding and discriminative relation scoring to generate and align relations effectively and efficiently during inference. Extensive experiments under fully supervised settings and low-resource settings demonstrate the effectiveness of our approach.
Fine-tuning pre-trained language models is a common practice in building NLP models for various tasks, including the case with less supervision. We argue that under the few-shot setting, formulating fine-tuning closer to the pre-training objective shall be able to unleash more benefits from the pre-trained language models. In this work, we take few-shot named entity recognition (NER) for a pilot study, where existing fine-tuning strategies are much different from pre-training. We propose a novel few-shot fine-tuning framework for NER, FFF-NER. Specifically, we introduce three new types of tokens, “is-entity”, “which-type” and “bracket”, so we can formulate the NER fine-tuning as (masked) token prediction or generation, depending on the choice of the pre-training objective. In our experiments, we apply to fine-tune both BERT and BART for few-shot NER on several benchmark datasets and observe significant improvements over existing fine-tuning strategies, including sequence labeling, prototype meta-learning, and prompt-based approaches. We further perform a series of ablation studies, showing few-shot NER performance is strongly correlated with the similarity between fine-tuning and pre-training.
We propose a simple ranking strategy to solve a generative commonsense question answering (QA) problem. Compared with multiple-choice QA, it is challenging because the answers to a question are not unique and they are supposed to be popular and diverse. Our strategy exploits the dataset itself and negative samples that we collect from WordNet to train a ranker that picks out the most popular answers for commonsense questions. The effectiveness of our strategy is verified on different pre-trained masked language models (MLMs) in a pipeline framework, where an MLM reranks the generated answers. Further, we explore an end-to-end framework where MLMs are utilized to guide the generation of generative language models (GLMs). Taking advantage of reinforcement learning, we apply policy gradient to train a GLM with the rewards fed back by an MLM. Empirical results on ProtoQA dataset demonstrate that MLMs can acquire the ability to distinguish the popular answers and improve the typical answer generation of GLMs as well.
While interacting with chatbots, users may elicit multiple intents in a single dialogue utterance. Instead of training a dedicated multi-intent detection model, we propose DialogUSR, a dialogue utterance splitting and reformulation task that first splits multi-intent user query into several single-intent sub-queries and then recovers all the coreferred and omitted information in the sub-queries. DialogUSR can serve as a plug-in and domain-agnostic module that empowers the multi-intent detection for the deployed chatbots with minimal efforts. We collect a high-quality naturally occurring dataset that covers 23 domains with a multi-step crowd-souring procedure. To benchmark the proposed dataset, we propose multiple action-based generative models that involve end-to-end and two-stage training, and conduct in-depth analyses on the pros and cons of the proposed baselines.
Recently, active learning (AL) methods have been used to effectively fine-tune pre-trained language models for various NLP tasks such as sentiment analysis and document classification. However, given the task of fine-tuning language models, understanding the impact of different aspects on AL methods such as labeling cost, sample acquisition latency, and the diversity of the datasets necessitates a deeper investigation. This paper examines the performance of existing AL methods within a low-resource, interactive labeling setting. We observe that existing methods often underperform in such a setting while exhibiting higher latency and a lack of generalizability. To overcome these challenges, we propose a novel active learning method TYROUGE that employs a hybrid sampling strategy to minimize labeling cost and acquisition latency while providing a framework for adapting to dataset diversity via user guidance. Through our experiments, we observe that compared to SOTA methods, TYROUGE reduces the labeling cost by up to 43% and the acquisition latency by as much as 11X, while achieving comparable accuracy. Finally, we discuss the strengths and weaknesses of TYROUGE by exploring the impact of dataset characteristics.
Simile recognition involves two subtasks: simile sentence classification that discriminates whether a sentence contains simile, and simile component extraction that locates the corresponding objects (i.e., tenors and vehicles).Recent work ignores features other than surface strings and suffers from the data hunger issue.We explore expressive features for this task to help achieve more effective data utilization.In particular, we study two types of features: 1) input-side features that include POS tags, dependency trees and word definitions, and 2) decoding features that capture the interdependence among various decoding decisions.We further construct a model named HGSR, which merges the input-side features as a heterogeneous graph and leverages decoding features via distillation.Experiments show that HGSR significantly outperforms the current state-of-the-art systems and carefully designed baselines, verifying the effectiveness of introduced features. We will release our code upon paper acceptance.
We propose a unified framework to generate both homophonic and homographic puns to resolve the split-up in existing works. Specifically, we incorporate three linguistic attributes of puns to the language models: ambiguity, distinctiveness, and surprise. Our framework consists of three parts: 1) a context words/phrases selector to promote the aforementioned attributes, 2) a generation model trained on non-pun sentences to incorporate the context words/phrases into the generation output, and 3) a label predictor that learns the structure of puns which is used to steer the generation model at inference time. Evaluation results on both pun types demonstrate the efficacy of our model over strong baselines.
Transliteration is an important task in natural language processing (NLP) which aims to convert a name in the source language to the target language without changing its pronunciation. Particularly, transliteration from English to Arabic is highly needed in many applications, especially in countries (e.g., United Arab Emirates (UAE)) whose most citizens are foreigners but the official language is Arabic. In such a task-oriented scenario, namely transliterating the English names to the corresponding Arabic ones, the performance of the transliteration model is highly important. However, most existing neural approaches mainly apply a universal transliteration model with advanced encoders and decoders to the task, where limited attention is paid to leveraging the phonemic association between English and Arabic to further improve model performance. In this paper, we focus on transliteration of people’s names from English to Arabic for the general public. In doing so, we collect a corpus named EANames by extracting high quality name pairs from online resources which better represent the names in the general public than linked Wikipedia entries that are always names of famous people). We propose a model for English-Arabic transliteration, where a memory module modeling the phonemic association between English and Arabic is used to guide the transliteration process. We run experiments on the collected data and the results demonstrate the effectiveness of our approach for English-Arabic transliteration.
Embedding-based approaches for dialog response retrieval embed the context-response pairs as points in the embedding space. These approaches are scalable, but fail to account for the complex, many-to-many relationships that exist between context-response pairs. On the other end of the spectrum, there are approaches that feed the context-response pairs jointly through multiple layers of neural networks. These approaches can model the complex relationships between context-response pairs, but fail to scale when the set of responses is moderately large (>1000). In this paper, we propose a scalable model that can learn complex relationships between context-response pairs. Specifically, the model maps the contexts as well as responses to probability distributions over the embedding space. We train the models by optimizing the Kullback-Leibler divergence between the distributions induced by context-response pairs in the training data. We show that the resultant model achieves better performance as compared to other embedding-based approaches on publicly available conversation data.
There are growing interests in adapting large-scale language models using parameter-efficient fine-tuning methods. However, accelerating the model itself and achieving better inference efficiency through model compression has not been thoroughly explored yet.Model compression could provide the benefits of reducing memory footprints, enabling low-precision computations, and ultimately achieving cost-effective inference.To combine parameter-efficient adaptation and model compression, we propose AlphaTuning consisting of post-training quantization of the pre-trained language model and fine-tuning only some parts of quantized parameters for a target task.Specifically, AlphaTuning works by employing binary-coding quantization, which factorizes the full-precision parameters into binary parameters and a separate set of scaling factors.During the adaptation phase, the binary values are frozen for all tasks, while the scaling factors are fine-tuned for the downstream task.We demonstrate that AlphaTuning, when applied to GPT-2 and OPT, performs competitively with full fine-tuning on a variety of downstream tasks while achieving >10x compression ratio under 4-bit quantization and >1,000x reduction in the number of trainable parameters.
The prosperity of Pretrained Language Models(PLM) has greatly promoted the development of Machine Reading Comprehension (MRC). However, these models are vulnerable and not robust to adversarial examples. In this paper, we propose Stable and Contrastive Question Answering (SCQA) to improve invariance of representation to alleviate these robustness issues. Specifically, we first construct positive example pairs which have same answer through data augmentation. Then SCQA learns enhanced representations with better alignment between positive pairs by introducing stability and contrastive loss. Experimental results show that our approach can boost the robustness of QA models cross different MRC tasks and attack sets significantly and consistently.
By explaining how humans would solve a given task, human rationales can provide strong learning signal for neural language models (NLMs). Explanation regularization (ER) aims to improve NLM generalization by pushing the NLM’s machine rationales (Which input tokens did the NLM focus on?) to align with human rationales (Which input tokens would humans focus on). Though prior works primarily study ER via in-distribution (ID) evaluation, out-of-distribution (OOD) generalization is often more critical in real-world scenarios, yet ER’s effect on OOD generalization has been underexplored.In this paper, we introduce ER-Test, a framework for evaluating ER models’ OOD generalization along three dimensions: unseen datasets, contrast set tests, and functional tests. Using ER-Test, we comprehensively analyze how ER models’ OOD generalization varies with the rationale alignment criterion (loss function), human rationale type (instance-level v/s task-level), number and choice of rationale-annotated instances, and time budget for rationale annotation. Across two tasks and six datasets, we show that ER has little impact on ID performance but yields large OOD performance gains, with the best ER criterion being task-dependent. Also, ER can improve OOD performance even with task-level or few human rationales. Finally, we find that rationale annotation is more time-efficient than label annotation for improving OOD performance. Our results with ER-Test help demonstrate ER’s utility and establish best practices for using ER effectively.
Aspect sentiment triplet extraction (ASTE) is an essential task, which aims to extract triplets(aspect, opinion, sentiment). However, overlapped triplets, especially multi-overlap triplets,make ASTE a challenge. Most existing methods suffer from multi-overlap triplets becausethey focus on the single interactions between an aspect and an opinion. To solve the aboveissues, we propose a novel multi-overlap triplet extraction method, which decodes the complexrelations between multiple aspects and opinions by learning their cooperative interactions. Overall, the method is based on an encoder-decoder architecture. During decoding, we design ajoint decoding mechanism, which employs a multi-channel strategy to generate aspects andopinions through the cooperative interactions between them jointly. Furthermore, we constructa correlation-enhanced network to reinforce the interactions between related aspectsand opinions for sentiment prediction. Besides, a relation-wise calibration scheme is adoptedto further improve performance. Experiments show that our method outperforms baselines,especially multi-overlap triplets.
Delta tuning (DET, also known as parameter-efficient tuning) is deemed as the new paradigm for using pre-trained language models (PLMs). Up to now, various DETs with distinct design elements have been proposed, achieving performance on par with fine-tuning. However, the mechanisms behind the above success are still under-explored, especially the connections among various DETs. To fathom the mystery, we hypothesize that the adaptations of different DETs could all be reparameterized as low-dimensional optimizations in a unified optimization subspace, which could be found by jointly decomposing independent solutions of different DETs. Then we explore the connections among different DETs by conducting optimization within the subspace. In experiments, we find that, for a certain DET, conducting optimization simply in the subspace could achieve comparable performance to its original space, and the found solution in the subspace could be transferred to another DET and achieve non-trivial performance. We also visualize the performance landscape of the subspace, and find that, there exists a substantial region where different DETs all perform well. Finally, we extend our analysis and show the strong connections between fine-tuning and DETs. The codes are publicly available at https://github.com/thunlp/Unified-DeltaTuning.
Joint intent detection and slot filling is a key research topic in natural language understanding (NLU). Existing joint intent and slot filling systems analyze and compute features collectively for all slot types, and importantly, have no way to explain the slot filling model decisions. In this work, we propose a novel approach that: (i) learns to generate additional slot type specific features in order to improve accuracy and (ii) provides explanations for slot filling decisions for the first time in a joint NLU model. We perform an additional constrained supervision using a set of binary classifiers for the slot type specific feature learning, thus ensuring appropriate attention weights are learned in the process to explain slot filling decisions for utterances. Our model is inherently explainable and does not need any post-hoc processing. We evaluate our approach on two widely used datasets and show accuracy improvements. Moreover, a detailed analysis is also provided for the exclusive slot explainability.
Commonsense Knowledge Base (CSKB) Population aims at reasoning over unseen entities and assertions on CSKBs, and is an important yet hard commonsense reasoning task. One challenge is that it requires out-of-domain generalization ability as the source CSKB for training is of a relatively smaller scale (1M) while the whole candidate space for population is way larger (200M). We propose PseudoReasoner, a semi-supervised learning framework for CSKB population that uses a teacher model pre-trained on CSKBs to provide pseudo labels on the unlabeled candidate dataset for a student model to learn from. The teacher can be a generative model rather than restricted to discriminative models as previous works.In addition, we design a new filtering procedure for pseudo labels based on influence function and the student model’s prediction to further improve the performance. The framework can improve the backbone model KG-BERT (RoBERTa-large) by 3.3 points on the overall performance and especially, 5.3 points on the out-of-domain performance, and achieves the state-of-the-art. The codes will be made public on acceptance. Codes and data are available at https://github.com/HKUST-KnowComp/PseudoReasoner.
With the evolution of pre-trained language models, current open-domain dialogue systems have achieved great progress in conducting one-session conversations. In contrast, Multi-Session Conversation (MSC), which consists of multiple sessions over a long term with the same user, is under-investigated. In this paper, we propose History-Aware Hierarchical Transformer (HAHT) for multi-session open-domain dialogue. HAHT maintains a long-term memory of history conversations and utilizes history information to understand current conversation context and generate well-informed and context-relevant responses. Specifically, HAHT first encodes history conversation sessions hierarchically into a history memory. Then, HAHT leverages historical information to facilitate the understanding of the current conversation context by encoding the history memory together with the current context with attention-based mechanisms. Finally, to explicitly utilize historical information, HAHT uses a history-aware response generator that switches between a generic vocabulary and a history-aware vocabulary. Experimental results on a large-scale MSC dataset suggest that the proposed HAHT model consistently outperforms baseline models. Human evaluation results support that HAHT generates more human-like, context-relevant, and history-relevant responses than baseline models.
Abstractive dialogue summarization has recently been receiving more attention. We propose a coarse-to-fine model for generating abstractive dialogue summaries, and introduce a fact-aware reinforcement learning (RL) objective that improves the fact consistency between the dialogue and the generated summary. Initially, the model generates the predicate-argument spans of the dialogue, and then generates the final summary through a fact-aware RL objective. Extensive experiments and analysis on two benchmark datasets demonstrate that our proposed method effectively improves the quality of the generated summary, especially in coherence and consistency.
Long samples of text from neural language models can be of poor quality. Truncation sampling algorithms–like top-p or top-k—address this by setting some words’ probabilities to zero at each step. This work investigates why these methods are important, and how to improve them. We propose thinking of a neural language model as a mixture of a true distribution and a smoothing distribution that avoids infinite perplexity. In this light, truncation algorithms aim to perform desmoothing, estimating a subset of the support of the true distribution. Finding a good subset is crucial: we show that top-p unnecessarily truncates high-probability words, for example causing it to truncate all words but Trump for a document that starts with Donald. We introduce eta-sampling, which truncates words below an entropy-dependent probability threshold. Compared to previous algorithms, our eta-sampling generates more plausible long documents according to humans, is better at breaking out of repetition, and behaves more reasonably on a battery of test distributions.
Knowledge (including structured knowledge such as schema and ontology and unstructured knowledge such as web corpus) is a critical part of dialog understanding, especially for unseen tasks and domains. Traditionally, such domain-specific knowledge is encoded implicitly into model parameters for the execution of downstream tasks, which makes training inefficient. In addition , such models are not easily transferable to new tasks with different schemas. In this work, we propose to perform dialog state tracking grounded on knowledge encoded externally. We query relevant knowledge of various forms based on the dialog context where such information can grounds the prediction of dialog states. We demonstrate superior performance of our proposed method over strong baselines, especially in the few-shot learning setting.
Supervised training of existing deep learning models for sequence labeling relies on large scale labeled datasets. Such datasets are generally created with crowd-source labeling. However, crowd-source labeling for tasks of sequence labeling can be expensive and time-consuming. Further, crowd-source labeling by external annotators may not be appropriate for data that contains user private information. Considering the above limitations of crowd-source labeling, we study interactive sequence labeling that allows training directly with the user feedback, which alleviates the annotation cost and maintains the user privacy. We identify two bias, namely, context bias and feedback bias, by formulating interactive sequence labeling via a Structural Causal Model (SCM). To alleviate the context and feedback bias based on the SCM, we identify the frequent context tokens as confounders in the backdoor adjustment and further propose an entropy-based modulation that is inspired by information theory. entities more sample-efficiently. With extensive experiments, we validate that our approach can effectively alleviate the biases and our models can be efficiently learnt with the user feedback.
Natural language inference (NLI) is a task to infer the relationship between a premise and a hypothesis (e.g., entailment, neutral, or contradiction), and transformer-based models perform well on current NLI datasets such as MNLI and SNLI. Nevertheless, given the linguistic complexity of the large-scale datasets, it remains controversial whether these models can truly infer the relationship between sentences or they simply guess the answer via shallow heuristics. Here, we introduce a controlled evaluation set called Simple Pair to test the basic sentence inference ability of NLI models using sentences with syntactically simple structures. Three popular transformer-based models, i.e., BERT, RoBERTa, and DeBERTa, are employed. We find that these models fine-tuned on MNLI or SNLI perform very poorly on Simple Pair (< 35.4% accuracy). Further analyses reveal event coreference and compositional binding problems in these models. To improve the model performance, we augment the training set, i.e., MNLI or SNLI, with a few examples constructed based on Simple Pair ( 1% of the size of the original SNLI/MNLI training sets). Models fine-tuned on the augmented training set maintain high performance on MNLI/SNLI and perform very well on Simple Pair (~100% accuracy). Furthermore, the positive performance of the augmented training models can transfer to more complex examples constructed based on sentences from MNLI and SNLI. Taken together, the current work shows that (1) models achieving high accuracy on mainstream large-scale datasets still lack the capacity to draw accurate inferences on simple sentences, and (2) augmenting mainstream datasets with a small number of target simple sentences can effectively improve model performance.
In recent years, there is a surge of generation-based information extraction work, which allows a more direct use of pre-trained language models and efficiently captures output dependencies. However, previous generative methods using lexical representation do not naturally fit document-level relation extraction (DocRE) where there are multiple entities and relational facts. In this paper, we investigate the root cause of the underwhelming performance of the existing generative DocRE models and discover that the culprit is the inadequacy of the training paradigm, instead of the capacities of the models. We propose to generate a symbolic and ordered sequence from the relation matrix which is deterministic and easier for model to learn. Moreover, we design a parallel row generation method to process overlong target sequences. Besides, we introduce several negative sampling strategies to improve the performance with balanced signals. Experimental results on four datasets show that our proposed method can improve the performance of the generative DocRE models.
Event argument extraction is a challenging subtask of event extraction, aiming to identify and assign roles to arguments under a certain event. Existing methods extract arguments of each role independently, ignoring the relationship between different roles. Such an approach hinders the model from learning explicit interactions between different roles to improve the performance of individual argument extraction. As a solution, we design a neural model that we refer to as the Explicit Role Interaction Network (ERIN) which allows for dynamically capturing the correlations between different argument roles within an event. Extensive experiments on the benchmark dataset ACE2005 demonstrate the superiority of our proposed model to existing approaches.
Training a model to provide natural language explanations (NLEs) for its predictions usually requires the acquisition of task-specific NLEs, which is time- and resource-consuming. A potential solution is the few-shot out-of-domain transfer of NLEs from a parent task with many NLEs to a child task.In this work, we examine the setup in which the child task has few NLEs but abundant labels. We establish four few-shot transfer learning methods that cover the possible fine-tuning combinations of the labels and NLEs for the parent and child tasks. We transfer explainability from a large natural language inference dataset (e-SNLI) separately to two child tasks: (1) hard cases of pronoun resolution, where we introduce the small-e-WinoGrande dataset of NLEs on top of the WinoGrande dataset, and (2) commonsense validation (ComVE). Our results demonstrate that the parent task helps with NLE generation and we establish the best methods for this setup.
Despite of the superb performance on a wide range of tasks, pre-trained language models (e.g., BERT) have been proved vulnerable to adversarial texts. In this paper, we present RoChBERT, a framework to build more Robust BERT-based models by utilizing a more comprehensive adversarial graph to fuse Chinese phonetic and glyph features into pre-trained representations during fine-tuning. Inspired by curriculum learning, we further propose to augment the training dataset with adversarial texts in combination with intermediate samples. Extensive experiments demonstrate that RoChBERT outperforms previous methods in significant ways: (i) robust – RoChBERT greatly improves the model robustness without sacrificing accuracy on benign texts. Specifically, the defense lowers the success rates of unlimited and limited attacks by 59.43% and 39.33% respectively, while remaining accuracy of 93.30%; (ii) flexible – RoChBERT can easily extend to various language models to solve different downstream tasks with excellent performance; and (iii) efficient – RoChBERT can be directly applied to the fine-tuning stage without pre-training language model from scratch, and the proposed data augmentation method is also low-cost.
In this paper, we introduce a novel method for lexical entailment tasks, which detects a hyponym-hypernym relation among words. Existing lexical entailment studies are lacking in generalization performance, as they cannot be applied to words that are not included in the training dataset. Moreover, existing work evaluates the performance by using the dataset that contains words used for training. This study proposes a method that learns a mapping from word embeddings to the hierarchical embeddings in order to predict the hypernymy relations of any input words. To validate the generalization performance, we conduct experiments using a train dataset that does not overlap with the evaluation dataset. As a result, our method achieved state-of-the-art performance and showed robustness for unknown words.
Prompt tuning, or the conditioning of a frozen pretrained language model (PLM) with soft prompts learned from data, has demonstrated impressive performance on a wide range of NLP tasks. However, prompt tuning requires a large training dataset to be effective and is outperformed by finetuning the entire PLM in data-scarce regimes. Previous work (Gu et al., 2022, Vu et al., 2022) proposed to transfer soft prompts pretrained on the source domain to the target domain. In this paper, we explore domain adaptation for prompt tuning, a problem setting where unlabeled data from the target domain are available during pretraining. We propose bOosting Prompt TunIng with doMain Adaptation (OPTIMA), which regularizes the decision boundary to be smooth around regions where source and target data distributions are similar. Extensive experiments demonstrate that OPTIMA significantly enhances the transferability and sample-efficiency of prompt tuning compared to strong baselines. Moreover, in few-shot settings, OPTIMA exceeds full-model tuning by a large margin.
Phrase similarity is a key component of many NLP applications. Current phrase similarity methods focus on embedding the phrase itself and use the phrase context only during training of the pretrained model. To better leverage the information in the context, we propose McPhraSy (Multi-context Phrase Similarity), a novel algorithm for estimating the similarity of phrases based on multiple contexts. At inference time, McPhraSy represents each phrase by considering multiple contexts in which it appears and computes the similarity of two phrases by aggregating the pairwise similarities between the contexts of the phrases. Incorporating context during inference enables McPhraSy to outperform current state-of-the-art models on two phrase similarity datasets by up to 13.3%. Finally, we also present a new downstream task that relies on phrase similarity – keyphrase clustering – and create a new benchmark for it in the product reviews domain. We show that McPhraSy surpasses all other baselines for this task.
Narrative modelling is an area of active research, motivated by the acknowledgement of narratives as drivers of societal decision making. These research efforts conceptualize narratives as connected entity chains, and modeling typically focuses on the identification of entities and their connections within a text. An emerging approach to narrative modelling is the use of semantic role labeling (SRL) to extract Entity-Verb-Entity (E-V-Es) tuples from a text, followed by dimensionality reduction to reduce the space of entities and connections separately. This process penalises the semantic richness of narratives and discards much contextual information along the way. Here, we propose an alternate narrative extraction approach - CANarEx, incorporating a pipeline of common contextual constructs through co-reference resolution, micro-narrative generation and clustering of these narratives through sentence embeddings. We evaluate our approach through testing the recovery of “narrative time-series clusters”, mimicking a desirable text-as-data task. The evaluation framework leverages synthetic data generated using a GPT-3 model. The GPT-3 model is trained to generate similar sentences using a large dataset of news articles. The synthetic data maps to three topics in the news dataset. We then generate narrative time-series document cluster representations by mapping the synthetic data to three distinct signals synthetically injected into the testing corpus. Evaluation results demonstrate the superior ability of CANarEx to recover narrative time-series through reduced MSE and improved precision/recall relative to existing methods. The validity is further reinforced through ablation studies and qualitative analysis.
Dialogue summarization models aim to generate a concise and accurate summary for multi-party dialogue. The complexity of dialogue, including coreference, dialogue acts, and inter-speaker interactions bring unique challenges to dialogue summarization. Most recent neural models achieve state-of-art performance following the pretrain-then-finetune recipe, where the large-scale language model (LLM) is pretrained on large-scale single-speaker written text, but later finetuned on multi-speaker dialogue text. To mitigate the gap between pretraining and finetuning, we propose several approaches to convert the dialogue into a third-person narrative style and show that the narration serves as a valuable annotation for LLMs. Empirical results on three benchmark datasets show our simple approach achieves higher scores on the ROUGE and a factual correctness metric.
Among all the safety concerns that hinder the deployment of open-domain dialog systems (e.g., offensive languages, biases, and toxic behaviors), social bias presents an insidious challenge. Addressing this challenge requires rigorous analyses and normative reasoning. In this paper, we focus our investigation on social bias measurement to facilitate the development of unbiased dialog systems. We first propose a novel Dial-Bias Framework for analyzing the social bias in conversations using a holistic method beyond bias lexicons or dichotomous annotations. Leveraging the proposed framework, we further introduce the CDial-Bias Dataset which is, to the best of our knowledge, the first annotated Chinese social bias dialog dataset. We also establish a fine-grained dialog bias measurement benchmark and conduct in-depth ablation studies to shed light on the utility of the detailed annotations in the proposed dataset. Finally, we evaluate representative Chinese generative models with our classifiers to unveil the presence of social bias in these systems.
Relation Extraction (RE) has attracted increasing attention, but current RE evaluation is limited to in-domain evaluation setups. Little is known on how well a RE system fares in challenging, but realistic out-of-distribution evaluation setups. To address this gap, we propose CrossRE, a new, freely-available cross-domain benchmark for RE, which comprises six distinct text domains and includes multi-label annotations. An additional innovation is that we release meta-data collected during annotation, to include explanations and flags of difficult instances. We provide an empirical evaluation with a state-of-the-art model for relation classification. As the meta-data enables us to shed new light on the state-of-the-art model, we provide a comprehensive analysis on the impact of difficult cases and find correlations between model and human annotations. Overall, our empirical investigation highlights the difficulty of cross-domain RE. We release our dataset, to spur more research in this direction.
Extracting fine-grained structural information between argumentation component (AC) pairs is essential for argumentation relation classification (ARC). However, most previous studies attempt to model the relationship between AC pairs using AC level similarity or semantically relevant features. They ignore the complex interaction between AC pairs and cannot effectively reason the argumentation relation deeply.Therefore, in this paper, we propose a novel dual prior graph neural network (DPGNN) to jointly explore the probing knowledge derived from pre-trained language models (PLMs) and the syntactical information for comprehensively modeling the relationship between AC pairs. Specifically, we construct a probing graph by using probing knowledge derived from PLMs to recognize and align the relational information within and across the argumentation components. In addition, we propose a mutual dependency graph for the AC pair to reason the fine-grained syntactic structural information, in which the syntactical correlation between words is set by the dependency information within AC and mutual attention mechanism across ACs. The knowledge learned from the probing graph and the dependency graph are combined to comprehensively capture the aligned relationships of AC pairs for improving the results of ARC. Experimental results on three public datasets show that DPGNN outperforms the state-of-the-art baselines by a noticeable margin.
The logical reasoning capabilities of pre-trained language models have recently received much attention. As one of the vital reasoning paradigms, non-monotonic reasoning refers to the fact that conclusions may be invalidated with new information. Existing work has constructed a non-monotonic inference dataset 𝛿-NLI and explored the performance of language models on it. However, the 𝛿-NLI dataset is entangled with commonsense reasoning. In this paper, we explore the pure non-monotonic reasoning ability of pre-trained language models. We build a non-monotonic reasoning benchmark, named LogicNMR, with explicit default rules and iterative updates. In the experimental part, the performance of popular language models on LogicNMR is explored from the perspectives of accuracy, generalization, proof-based traceability and robustness. The experimental results show that even though the fine-tuned language models achieve an accuracy of more than 94.4% on LogicNMR, they perform unsatisfactorily, with a significant drop, in generalization and proof-based traceability.
For humans and computers, the first step in answering an open-domain question is retrieving a set of relevant documents from a large corpus. However, the strategies that computers use fundamentally differ from those of humans. To better understand these differences, we design a gamified interface for data collection—Cheater’s Bowl—where a human answers complex questions with access to both traditional and modern search tools. We collect a dataset of human search sessions, analyze human search strategies, and compare them to state-of-the-art multi-hop QA models. Humans query logically, apply dynamic search chains, and use world knowledge to boost searching. We demonstrate how human queries can improve the accuracy of existing systems and propose improving the future design of QA models.
Despite being able to generate fluent and grammatical text, current Seq2Seq summarization models still suffering from the unfaithful generation problem.In this paper, we study the faithfulness of existing systems from a new perspective of factual robustness which is the ability to correctly generate factual information over adversarial unfaithful information.We first measure a model’sfactual robustness by its success rate to defend against adversarial attacks when generating factual information.The factual robustness analysis on a wide range of current systems shows its good consistency with human judgments on faithfulness.Inspired by these findings, we propose to improve the faithfulness of a model by enhancing its factual robustness.Specifically, we propose a novel training strategy, namely FRSUM, which teaches the model to defend against both explicit adversarial samples and implicit factual adversarial perturbations.Extensive automatic and human evaluation results show that FRSUM consistently improves the faithfulness of various Seq2Seq models, such as T5, BART.
Formal verse poetry imposes strict constraints on the meter and rhyme scheme of poems. Most prior work on generating this type of poetry uses existing poems for supervision, which are difficult to obtain for most languages and poetic forms. In this work, we propose an unsupervised approach to generate poems that follow any given meter and rhyme scheme, without requiring any poetic text for training. Our method works by splitting a regular, non-poetic corpus into phrases, prepending control codes that describe the length and end rhyme of each phrase, and training a transformer language model in the augmented corpus. The transformer learns to link the structure descriptor with the control codes to the number of lines, their length and their end rhyme. During inference, we build control codes for the desired meter and rhyme scheme, and condition our language model on them to generate formal verse poetry. Experiments in Spanish and Basque show that our approach is able to generate valid poems, which are often comparable in quality to those written by humans.
Recently, dataset-generation-based zero-shot learning has shown promising results by training a task-specific model with a dataset synthesized from large pre-trained language models (PLMs). The final task-specific model often achieves compatible or even better performance than PLMs under the zero-shot setting, with orders of magnitude fewer parameters.However, synthetic datasets have their drawbacks. They have long being suffering from the low-quality issue (e.g., low informativeness, redundancy). This explains why the massive synthetic data does not lead to better performance – a scenario we would expect in the human-labeled data. To improve the quality in dataset synthesis, we propose a progressive zero-shot dataset generation framework, ProGen, which leverages the feedback from the task-specific model to guide the generation of new training data via in-context examples.Extensive experiments on five text classification datasets demonstrate the effectiveness of the proposed approach. We also show ProGen achieves on-par or superior performance with only 1% synthetic dataset size, when comparing to baseline methods without in-context feedback.
Large pretrained language models can easily produce toxic or biased content, which is prohibitive for practical use. In order to detect such toxic generations, existing methods rely on templates, real-world data extraction, crowdsourcing workers or automatic generation to construct adversarial contexts that are likely to induce toxic generations. However, what type of context is more likely to induce unsafe responses is still under-explored. In this paper, we identify that context toxicity and context category (e.g., profanity, insult, drugs, etc.) are two important factors to cause safety issues in response generation. Hence, we propose a method called reverse generation to construct adversarial contexts conditioned on a given response, with the flexibility to control category, toxicity level and inductivity of the generated contexts. Via reverse generation, we augment the existing BAD dataset and construct a new dataset BAD+ which contains more than 120K diverse and highly inductive contexts in 12 categories. We test three popular pretrained dialogue models (Blender, DialoGPT and Plato2) and find that BAD+ can largely expose their safety problems. Furthermore, we show that BAD+ can greatly enhance the safety of generation, and we reveal the key factors of safety improvement. Our code and dataset is available at https://github.com/thu-coai/Reverse_Generation.
Visual Question Answering (VQA) models are prone to learn the shortcut solution formed by dataset biases rather than the intended solution. To evaluate the VQA models’ reasoning ability beyond shortcut learning, the VQA-CP v2 dataset introduces a distribution shift between the training and test set given a question type. In this way, the model cannot use the training set shortcut (from question type to answer) to perform well on the test set. However, VQA-CP v2 only considers one type of shortcut and thus still cannot guarantee that the model relies on the intended solution rather than a solution specific to this shortcut. To overcome this limitation, we propose a new dataset that considers varying types of shortcuts by constructing different distribution shifts in multiple OOD test sets. In addition, we overcome the three troubling practices in the use of VQA-CP v2, e.g., selecting models using OOD test sets, and further standardize OOD evaluation procedure. Our benchmark provides a more rigorous and comprehensive testbed for shortcut learning in VQA. We benchmark recent methods and find that methods specifically designed for particular shortcuts fail to simultaneously generalize to our varying OOD test sets. We also systematically study the varying shortcuts and provide several valuable findings, which may promote the exploration of shortcut learning in VQA.
Building dense retrievers requires a series of standard procedures, including training and validating neural models and creating indexes for efficient search. However, these procedures are often misaligned in that training objectives do not exactly reflect the retrieval scenario at inference time. In this paper, we explore how the gap between training and inference in dense retrieval can be reduced, focusing on dense phrase retrieval (Lee et al., 2021) where billions of representations are indexed at inference. Since validating every dense retriever with a large-scale index is practically infeasible, we propose an efficient way of validating dense retrievers using a small subset of the entire corpus. This allows us to validate various training strategies including unifying contrastive loss terms and using hard negatives for phrase retrieval, which largely reduces the training-inference discrepancy. As a result, we improve top-1 phrase retrieval accuracy by 2 3 points and top-20 passage retrieval accuracy by 2 4 points for open-domain question answering. Our work urges modeling dense retrievers with careful consideration of training and inference via efficient validation while advancing phrase retrieval as a general solution for dense retrieval.
While the NLP community is generally aware of resource disparities among languages, we lack research that quantifies the extent and types of such disparity. Prior surveys estimating the availability of resources based on the number of datasets can be misleading as dataset quality varies: many datasets are automatically induced or translated from English data. To provide a more comprehensive picture of language resources, we examine the characteristics of 156 publicly available NLP datasets. We manually annotate how they are created, including input text and label sources and tools used to build them, and what they study, tasks they address and motivations for their creation. After quantifying the qualitative NLP resource gap across languages, we discuss how to improve data collection in low-resource languages. We survey language-proficient NLP researchers and crowd workers per language, finding that their estimated availability correlates with dataset availability. Through crowdsourcing experiments, we identify strategies for collecting high-quality multilingual data on the Mechanical Turk platform. We conclude by making macro and micro-level suggestions to the NLP community and individual researchers for future multilingual data development.
Recent years have witnessed the rise and success of pre-training techniques in visually-rich document understanding. However, most existing methods lack the systematic mining and utilization of layout-centered knowledge, leading to sub-optimal performances. In this paper, we propose ERNIE-Layout, a novel document pre-training solution with layout knowledge enhancement in the whole workflow, to learn better representations that combine the features from text, layout, and image. Specifically, we first rearrange input sequences in the serialization stage, and then present a correlative pre-training task, reading order prediction, to learn the proper reading order of documents. To improve the layout awareness of the model, we integrate a spatial-aware disentangled attention into the multi-modal transformer and a replaced regions prediction task into the pre-training phase. Experimental results show that ERNIE-Layout achieves superior performance on various downstream tasks, setting new state-of-the-art on key information extraction, document image classification, and document question answering datasets. The code and models are publicly available at PaddleNLP.
The charge prediction task aims to predict the charge for a case given its fact description. Recent models have already achieved impressive accuracy in this task, however, little is understood about the mechanisms they use to perform the judgment.For practical applications, a charge prediction model should conform to the certain legal theory in civil law countries, as under the framework of civil law, all cases are judged according to certain local legal theories. In China, for example, nearly all criminal judges make decisions based on the Four Elements Theory (FET).In this paper, we argue that trustworthy charge prediction models should take legal theories into consideration, and standing on prior studies in model interpretation, we propose three principles for trustworthy models should follow in this task, which are sensitive, selective, and presumption of innocence.We further design a new framework to evaluate whether existing charge prediction models learn legal theories. Our findings indicate that, while existing charge prediction models meet the selective principle on a benchmark dataset, most of them are still not sensitive enough and do not satisfy the presumption of innocence. Our code and dataset are released at https://github.com/ZhenweiAn/EXP_LJP.
Remembering important information from the past and continuing to talk about it in the present are crucial in long-term conversations. However, previous literature does not deal with cases where the memorized information is outdated, which may cause confusion in later conversations. To address this issue, we present a novel task and a corresponding dataset of memory management in long-term conversations, in which bots keep track of and bring up the latest information about users while conversing through multiple sessions. In order to support more precise and interpretable memory, we represent memory as unstructured text descriptions of key information and propose a new mechanism of memory management that selectively eliminates invalidated or redundant information. Experimental results show that our approach outperforms the baselines that leave the stored memory unchanged in terms of engagingness and humanness, with larger performance gap especially in the later sessions.
Pre-trained language models have shown superior performance in task-oriented dialogues. However, existing datasets are on limited scales, which cannot support large-scale pre-training. Fortunately, various data augmentation methods have been developed to augment large-scale task-oriented dialogue corpora. However, they heavily rely on annotated data in the target domain, which require a tremendous amount of data collection and human labeling work. In this paper, we build a unified dialogue user simulation model by pre-training on several publicly available datasets. The model can then be tuned on a target domain with few-shot data. The experiments on a target dataset across multiple domains show that our proposed model brings remarkable performance increases through data augmentation.
Although existing neural network approaches have achieved great progress on Chinese spelling correction, there is still room to improve. The model is required to avoid over-correction and to distinguish a correct token from its phonological and visual similar ones. In this paper, we propose an error-guided correction model to address these issues. By borrowing the powerful ability of the pre-trained BERT model, we propose a novel zero-shot error detection method to do a preliminary detection, which guides our model to attend more on the probably wrong tokens in encoding and to avoid modifying the correct tokens in generating. Furthermore, we introduce a new loss function to integrate the error confusion set, which enables our model to distinguish similar tokens. Moreover, our model supports highly parallel decoding to meet real applications. Experiments are conducted on widely used benchmarks. Our model achieves superior performance against state-of-the-art approaches by a remarkable margin, on both the quality and computation speed.
We seek to semantically describe a set of images, capturing both the attributes of single images and the variations within the set. Our procedure is analogous to Principle Component Analysis, in which the role of projection vectors is replaced with generated phrases. First, a centroid phrase that has the largest average semantic similarity to the images in the set is generated, where both the computation of the similarity and the generation are based on pretrained vision-language models. Then, the phrase that generates the highest variation among the similarity scores is generated, using the same models. The next phrase maximizes the variance subject to being orthogonal, in the latent space, to the highest-variance phrase, and the process continues. Our experiments show that our method is able to convincingly capture the essence of image sets and describe the individual elements in a semantically meaningful way within the context of the entire set. Our code is available at: https://github.com/OdedH/textual-pca.
Most existing sequence generation models produce outputs in one pass, usually left-to-right. However, this is in contrast with a more natural approach that humans use in generating content; iterative refinement and editing. Recent work has introduced edit-based models for various tasks (such as neural machine translation and text style transfer), but these generally model a single edit step. In this work, we propose modeling editing processes, modeling the whole process of iteratively generating sequences. We form a conceptual framework to describe the likelihood of multi-step edits, and describe neural models that can learn a generative model of sequences based on these multistep edits. We introduce baseline results and metrics on this task, finding that modeling editing processes improves performance on a variety of axes on both our proposed task and related downstream tasks compared to previous single-step models of edits.
This paper presents a parameter-lite transfer learning approach of pretrained language models (LM) for knowledge graph (KG) completion. Instead of finetuning, which modifies all LM parameters, we only tune a few new parameters while keeping the original LM parameters fixed. We establish this via reformulating KG completion as a “fill-in-the-blank” task, and introducing a parameter-lite encoder on top of the original LMs. We show that, by tuning far fewer parameters than finetuning, LMs transfer non-trivially to most tasks and reach competitiveness with prior state-of-the-art approaches. For instance, we outperform the fully finetuning approaches on a KG completion benchmark by tuning only 1% of the parameters.
Due to the absence of connectives, implicit discourse relation recognition (IDRR) is still a challenging and crucial task in discourse analysis. Most of the current work adopted multitask learning to aid IDRR through explicit discourse relation recognition (EDRR) or utilized dependencies between discourse relation labels to constrain model predictions. But these methods still performed poorly on fine-grained IDRR and even utterly misidentified on most of the few-shot discourse relation classes. To address these problems, we propose a novel Prompt-based Connective Prediction (PCP) method for IDRR. Our method instructs large-scale pre-trained models to use knowledge relevant to discourse relation and utilizes the strong correlation between connectives and discourse relation to help the model recognize implicit discourse relations. Experimental results show that our method surpasses the current state-of-the-art model and achieves significant improvements on those fine-grained few-shot discourse relation. Moreover, our approach is able to be transferred to EDRR and obtain acceptable results. Our code is released in https://github.com/zh-i9/PCP-for-IDRR.
Performance of downstream NLP tasks on code-switched Hindi-English (aka ) continues to remain a significant challenge. Intuitively, Hindi and English corpora should aid improve task performance on Hinglish. We show that meta-learning framework can effectively utilize the the labelled resources of the downstream tasks in the constituent languages. The proposed approach improves the performance on downstream tasks on code-switched language. We experiment with code-switching benchmark GLUECoS and report significant improvements.
Knowledge Base Question Answering (KBQA) involving complex reasoning is emerging as an important research direction. However, most KBQA systems struggle with generalizability, particularly on two dimensions: (a) across multiple knowledge bases, where existing KBQA approaches are typically tuned to a single knowledge base, and (b) across multiple reasoning types, where majority of datasets and systems have primarily focused on multi-hop reasoning. In this paper, we present SYGMA, a modular KBQA approach developed with goal of generalization across multiple knowledge bases and multiple reasoning types. To facilitate this, SYGMA is designed as two high level modules: 1) KB-agnostic question understanding module that remain common across KBs, and generates logic representation of the question with high level reasoning constructs that are extensible, and 2) KB-specific question mapping and answering module to address the KB-specific aspects of the answer extraction. We evaluated SYGMA on multiple datasets belonging to distinct knowledge bases (DBpedia and Wikidata) and distinct reasoning types (multi-hop and temporal). State-of-the-art or competitive performances achieved on those datasets demonstrate its generalization capability.
Few-shot text matching is a more practical technique in natural language processing (NLP) to determine whether two texts are semantically identical. They primarily design patterns to reformulate text matching into a pre-trained task with uniform prompts across all instances. But they fail to take into account the connection between prompts and instances. This paper argues that dynamically strengthening the correlation between particular instances and the prompts is necessary because fixed prompts cannot adequately fit all diverse instances in inference. We suggest IGATE: Instance-Guided prompt leArning for few-shoT tExt matching, a novel pluggable prompt learning method. The gate mechanism used by IGATE, which is between the embedding and the PLM encoders, makes use of the semantics of instances to regulate the effects of the gate on the prompt tokens. The experimental findings show that IGATE achieves SOTA performance on MRPC and QQP, outperforming strong baselines. GitHub will host the release of codes.
We present M3 (Multi-level dataset for Multi-document summarisation of Medical studies), a benchmark dataset for evaluating the quality of summarisation systems in the biomedical domain. The dataset contains sets of multiple input documents and target summaries of three levels of complexity: documents, sentences, and propositions. The dataset also includes several levels of annotation, including biomedical entities, direction, and strength of relations between them, and the discourse relationships between the input documents (“contradiction” or “agreement”). We showcase usage scenarios of the dataset by testing 10 generic and domain-specific summarisation models in a zero-shot setting, and introduce a probing task based on counterfactuals to test if models are aware of the direction and strength of the conclusions generated from input studies.
Large language models appear to learn facts from the large text corpora they are trained on. Such facts are encoded implicitly within their many parameters, making it difficult to verify or manipulate what knowledge has been learned. Language models have recently been extended to multilingual language models (MLLMs), enabling knowledge to be learned across hundreds of languages. Meanwhile, knowledge graphs contain facts in an explicit triple format, which require careful and costly curation and are only available in a few high-resource languages, restricting their research and application. To address these issues, we propose to enhance MLLMs with knowledge from multilingual knowledge graphs (MLKGs) so as to tackle language and knowledge graph tasks across many languages, including low-resource ones. Specifically, we introducea lightweight adapter set to enhance MLLMs with cross-lingual entity alignment and facts from MLKGs for many languages. Experiments on common benchmarks show that such enhancement benefits both MLLMs and MLKGs, achieving: (1) comparable or improved performance for knowledge graph completion and entity alignment relative to baselines, especially for low-resource languages (for which knowledge graphs are unavailable); and (2) improved MLLM performance on language understanding tasks that require multilingual factual knowledge; all while maintaining performance on other general language tasks.
In the weakly supervised learning paradigm, labeling functions automatically assign heuristic, often noisy, labels to data samples. In this work, we provide a method for learning from weak labels by separating two types of complementary information associated with the labeling functions: information related to the target label and information specific to one labeling function only. Both types of information are reflected to different degrees by all labeled instances. In contrast to previous works that aimed at correcting or removing wrongly labeled instances, we learn a branched deep model that uses all data as-is, but splits the labeling function information in the latent space. Specifically, we propose the end-to-end model SepLL which extends a transformer classifier by introducing a latent space for labeling function specific and task-specific information. The learning signal is only given by the labeling functions matches, no pre-processing or label model is required for our method. Notably, the task prediction is made from the latent layer without any direct task signal. Experiments on Wrench text classification tasks show that our model is competitive with the state-of-the-art, and yields a new best average performance.
Understanding relational knowledge plays an integral part in natural language comprehension. When it comes to pre-trained language models (PLM), prior work has been focusing on probing relational knowledge this by filling the blanks in pre-defined prompts such as “The capital of France is —". However, these probes may be affected by the co-occurrence of target relation words and entities (e.g. “capital”, “France” and “Paris”) in the pre-training corpus. In this work, we extend these probing methodologies leveraging analogical proportions as a proxy to probe relational knowledge in transformer-based PLMs without directly presenting the desired relation. In particular, we analysed the ability of PLMs to understand (1) the directionality of a given relation (e.g. Paris-France is not the same as France-Paris); (2) the ability to distinguish types on a given relation (both France and Japan are countries); and (3) the relation itself (Paris is the capital of France, but not Rome). Our results show how PLMs are extremely accurate at (1) and (2), but have clear room for improvement for (3). To better understand the reasons behind this behaviour and mistakes made by PLMs, we provide an extended quantitative analysis based on relevant factors such as frequency.
Lifelong learning aims to accumulate knowledge and alleviate catastrophic forgetting when learning tasks sequentially. However, existing lifelong language learning methods only focus on the supervised learning setting. Unlabeled data, which can be easily accessed in real-world scenarios, are underexplored. In this paper, we explore a novel setting, semi-supervised lifelong language learning (SSLL), where a model learns sequentially arriving language tasks with both labeled and unlabeled data. We propose an unlabeled data enhanced lifelong learner to explore SSLL. Specially, we dedicate task-specific modules to alleviate catastrophic forgetting and design two modules to exploit unlabeled data: (1) a virtual supervision enhanced task solver is constructed on a teacher-student framework to mine the underlying knowledge from unlabeled data; and (2) a backward augmented learner is built to encourage knowledge transfer from newly arrived unlabeled data to previous tasks. Experimental results on various language tasks demonstrate our model’s effectiveness and superiority over competitive baselines under the new setting SSLL.
Prompt-based learning has achieved excellent performance in few-shot learning by mapping the outputs of the pre-trained language model to the labels with the help of a label mapping component. Existing manual label mapping (MLM) methods achieve good results but heavily rely on expensive human knowledge. Automatic label mapping (ALM) methods that learn the mapping functions with extra parameters have shown their potentiality. However, no effective ALM model comparable to MLM methods is developed yet due to the limited data. In this paper, we propose a Latent Pseudo Label Mapping (LPLM) method that optimizes the label mapping without human knowledge and extra parameters. LPLM is built upon a probabilistic latent model and is iteratively self-improved with the EM-style algorithm. The empirical results demonstrate that our LPLM method is superior to the mainstream ALM methods and significantly outperforms the SOTA method in few-shot classification tasks. Moreover, LPLM also shows impressively better performance than the vanilla MLM method which requires extra task-specific prior knowledge.
In logographic languages like Chinese, word meanings are constructed using specific character formations, which can help to disambiguate word senses and are beneficial for sentiment classification. However, such knowledge is rarely explored in previous sentiment analysis methods. In this paper, we focus on exploring the logographic information for aspect-based sentiment classification in Chinese text. Specifically, we employ a logographic image to capture an internal morphological structure from the character sequence. The logographic image is also used to learn the external relations among context and aspect words. Furthermore, we propose a multimodal language model to explicitly incorporate a logographic image with review text for aspect-based sentiment classification in Chinese. Experimental results show that our method brings substantial performance improvement over strong baselines. The results also indicate that the logographic image is very important for exploring the internal structure and external relations from the character sequence.
Prior work on language model pre-training has explored different architectures and learning objectives, but differences in data, hyperparameters and evaluation make a principled comparison difficult. In this work, we focus on bidirectionality as a key factor that differentiates existing approaches, and present a comprehensive study of its role in next token prediction, text infilling, zero-shot priming and fine-tuning. We propose a new framework that generalizes prior approaches, including fully unidirectional models like GPT, fully bidirectional models like BERT, and hybrid models like CM3 and prefix LM. Our framework distinguishes between two notions of bidirectionality (bidirectional context and bidirectional attention) and allows us to control each of them separately. We find that the optimal configuration is largely application-dependent (e.g., bidirectional attention is beneficial for fine-tuning and infilling, but harmful for next token prediction and zero-shot priming). We train models with up to 6.7B parameters, and find differences to remain consistent at scale. While prior work on scaling has focused on left-to-right autoregressive models, our results suggest that this approach comes with some trade-offs, and it might be worthwhile to develop very large bidirectional models.
Expressing attitude or stance toward entities and concepts is an integral part of human behavior and personality. Recently, evaluative language data has become more accessible with social media’s rapid growth, enabling large-scale opinion analysis. However, surprisingly little research examines the relationship between personality and evaluative language. To bridge this gap, we introduce the notion of evaluative topics, obtained by applying topic models to pre-filtered evaluative text from social media. We then link evaluative topics to individual text authors to build their evaluative profiles. We apply evaluative profiling to Reddit comments labeled with personality scores and conduct an exploratory study on the relationship between evaluative topics and Big Five personality facets, aiming for a more interpretable, facet-level analysis. Finally, we validate our approach by observing correlations consistent with prior research in personality psychology.
Code pre-trained models (CodePTMs) have recently demonstrated significant success in code intelligence. To interpret these models, some probing methods have been applied. However, these methods fail to consider the inherent characteristics of codes. In this paper, to address the problem, we propose a novel probing method CAT-probing to quantitatively interpret how CodePTMs attend code structure. We first denoise the input code sequences based on the token types pre-defined by the compilers to filter those tokens whose attention scores are too small. After that, we define a new metric CAT-score to measure the commonality between the token-level attention scores generated in CodePTMs and the pair-wise distances between corresponding AST nodes. The higher the CAT-score, the stronger the ability of CodePTMs to capture code structure. We conduct extensive experiments to integrate CAT-probing with representative CodePTMs for different programming languages. Experimental results show the effectiveness of CAT-probing in CodePTM interpretation. Our codes and data are publicly available at https://github.com/nchen909/CodeAttention.
In real-world scenarios with naturally occurring datasets, reference summaries are noisy and may contain information that cannot be inferred from the source text. On large news corpora, removing low quality samples has been shown to reduce model hallucinations. Yet, for smaller, and/or noisier corpora, filtering is detrimental to performance. To improve reference quality while retaining all data, we propose a new approach: to selectively re-write unsupported reference sentences to better reflect source data. We automatically generate a synthetic dataset of positive and negative revisions by corrupting supported sentences and learn to revise reference sentences with contrastive learning. The intensity of revisions is treated as a controllable attribute so that, at inference, diverse candidates can be over-generated-then-rescored to balance faithfulness and abstraction. To test our methods, we extract noisy references from publicly available MIMIC-III discharge summaries for the task of hospital-course summarization, and vary the data on which models are trained. According to metrics and human evaluation, models trained on revised clinical references are much more faithful, informative, and fluent than models trained on original or filtered data.
Recent applications of natural language processing techniques to suicidal ideation detection and risk assessment frame the detection or assessment task as a text classification problem. Recent advances have developed many models, especially deep learning models, to boost predictive performance.Though the performance (in terms of aggregated evaluation scores) is improving, this position paper urges that better intention understanding is required for reliable suicidal risk assessment with computational methods. This paper reflects the state of natural language processing applied to suicide-associated text classification tasks, differentiates suicidal risk assessment and intention understanding, and points out potential limitations of sentiment features and pretrained language models in suicidal intention understanding.Besides, it urges the necessity for sequential intention understanding and risk assessment, discusses some critical issues in evaluation such as uncertainty, and studies the lack of benchmarks.
Explanation faithfulness of model predictions in natural language processing is typically evaluated on held-out data from the same temporal distribution as the training data (i.e. synchronous settings). While model performance often deteriorates due to temporal variation (i.e. temporal concept drift), it is currently unknown how explanation faithfulness is impacted when the time span of the target data is different from the data used to train the model (i.e. asynchronous settings). For this purpose, we examine the impact of temporal variation on model explanations extracted by eight feature attribution methods and three select-then-predict models across six text classification tasks. Our experiments show that (i) faithfulness is not consistent under temporal variations across feature attribution methods (e.g. it decreases or increases depending on the method), with an attention-based method demonstrating the most robust faithfulness scores across datasets; and (ii) select-then-predict models are mostly robust in asynchronous settings with only small degradation in predictive performance. Finally, feature attribution methods show conflicting behavior when used in FRESH (i.e. a select-and-predict model) and for measuring sufficiency/comprehensiveness (i.e. as post-hoc methods), suggesting that we need more robust metrics to evaluate post-hoc explanation faithfulness. Code will be made publicly available.
We consider the task of image-captioning using only the CLIP model and additional text data at training time and no additional captioned images. Our approach relies on the fact that CLIP is trained to make visual and textual embeddings similar. Therefore, we only need to learn how to translate CLIP textual embeddings back into text, and we can learn how to do this by learning a decoder for the frozen CLIP text encoder using only text. We argue that this intuition is “almost correct” because of a gap between the embedding spaces, and propose to rectify this via noise injection during training. We demonstrate the effectiveness of our approach by showing SOTA zero-shot image captioning across four benchmarks, including style transfer. Code, data, and models are available at https://github.com/DavidHuji/CapDec.
Fine-tuning large pretrained language models on a limited training corpus usually suffers from poor generalization. Prior works show that the recently-proposed sharpness-aware minimization (SAM) optimization method can improve the model generalization. However, SAM adds a perturbation to each model parameter equally (but not all parameters contribute equally to the optimization of training), which we argue is sub-optimal and will lead to excessive computation. In this paper, we propose a novel optimization procedure, namely FSAM, which introduces a Fisher mask to improve the efficiency and performance of SAM. In short, instead of adding perturbation to all parameters, FSAM uses the Fisher information to identity the important parameters and formulates a Fisher mask to obtain the sparse perturbation, i.e., making the optimizer focus on these important parameters. Experiments on various tasks in GLUE and SuperGLUE benchmarks show that FSAM consistently outperforms the vanilla SAM by 0.67 1.98 average score among four different pretrained models. We also empirically show that FSAM works well in other complex scenarios, e.g., fine-tuning on generation tasks or limited training data. Encouragingly, when training data is limited, FSAM improves the SAM by a large margin, i.e., up to 15.1.
Transformer-based language models achieve state-of-the-art results on several natural language processing tasks. One of these is textual entailment, i.e., the task of determining whether a premise logically entails a hypothesis. However, the models perform poorly on this task when the examples contain negations. In this paper, we propose a new definition of textual entailment that captures also negation. This allows us to develop TINA (Textual Inference with Negation Augmentation), a principled technique for negated data augmentation that can be combined with the unlikelihood loss function.Our experiments with different transformer-based models show that our method can significantly improve the performance of the models on textual entailment datasets with negation – without sacrificing performance on datasets without negation.
Bilingual lexicon induction (BLI) with limited bilingual supervision is a crucial yet challenging task in multilingual NLP. Current state-of-the-art BLI methods rely on the induction of cross-lingual word embeddings (CLWEs) to capture cross-lingual word similarities; such CLWEs are obtained <b>1)</b> via traditional static models (e.g., VecMap), or <b>2)</b> by extracting type-level CLWEs from multilingual pretrained language models (mPLMs), or <b>3)</b> through combining the former two options. In this work, we propose a novel semi-supervised <i>post-hoc</i> reranking method termed <b>BLICEr</b> (<b>BLI</b> with <b>C</b>ross-<b>E</b>ncoder <b>R</b>eranking), applicable to any precalculated CLWE space, which improves their BLI capability. The key idea is to ‘extract’ cross-lingual lexical knowledge from mPLMs, and then combine it with the original CLWEs. This crucial step is done via <b>1)</b> creating a word similarity dataset, comprising positive word pairs (i.e., true translations) and hard negative pairs induced from the original CLWE space, and then <b>2)</b> fine-tuning an mPLM (e.g., mBERT or XLM-R) in a cross-encoder manner to predict the similarity scores. At inference, we <b>3)</b> combine the similarity score from the original CLWE space with the score from the BLI-tuned cross-encoder. BLICEr establishes new state-of-the-art results on two standard BLI benchmarks spanning a wide spectrum of diverse languages: it substantially outperforms a series of strong baselines across the board. We also validate the robustness of BLICEr with different CLWEs.
Retrieving evidences from tabular and textual resources is essential for open-domain question answering (OpenQA), which provides more comprehensive information. However, training an effective dense table-text retriever is difficult due to the challenges of table-text discrepancy and data sparsity problem. To address the above challenges, we introduce an optimized OpenQA Table-Text Retriever (OTTeR) to jointly retrieve tabular and textual evidences. Firstly, we propose to enhance mixed-modality representation learning via two mechanisms: modality-enhanced representation and mixed-modality negative sampling strategy. Secondly, to alleviate data sparsity problem and enhance the general retrieval ability, we conduct retrieval-centric mixed-modality synthetic pre-training. Experimental results demonstrate that OTTeR substantially improves the performance of table-and-text retrieval on the OTT-QA dataset. Comprehensive analyses examine the effectiveness of all the proposed mechanisms. Besides, equipped with OTTeR, our OpenQA system achieves the state-of-the-art result on the downstream QA task, with 10.1% absolute improvement in terms of the exact match over the previous best system.
In an effort to study the inductive biases of language models, numerous studies have attempted to use linguistically motivated tasks as a proxy of sorts, wherein performance on these tasks would imply an inductive bias towards a specific linguistic phenomenon. In this study, we attempt to analyse the inductive biases of language models with respect to natural language phenomena, in the context of building multilingual embedding spaces.We sample corpora from 2 sources in 15 languages and train language models on pseudo-bilingual variants of each corpus, created by duplicating each corpus and shifting token indices for half the resulting corpus. We evaluate the cross-lingual capabilities of these LMs, and show that while correlations with language families tend to be weak, other corpus-level characteristics, such as type-token ratio, tend to be more strongly correlated. Finally, we show that multilingual spaces can be built, albeit less effectively, even when additional destructive perturbations are applied to the training corpora, implying that (effectively) bag-of-words models also have an inductive bias that is sufficient for inducing multilingual spaces.
Complex dialogue mappings (CDM), including one-to-many and many-to-one mappings, tend to make dialogue models generate incoherent or dull responses, and modeling these mappings remains a huge challenge for neural dialogue systems. To alleviate these problems, methods like introducing external information, reconstructing the optimization function, and manipulating data samples are proposed, while they primarily focus on avoiding training with CDM, inevitably weakening the model’s ability of understanding CDM in human conversations and limiting further improvements in model performance. This paper proposes a Sentence Semantic Segmentation guided Conditional Variational Auto-Encoder (SegCVAE) method which can model and take advantages of the CDM data. Specifically, to tackle the incoherent problem caused by one-to-many, SegCVAE uses response-related prominent semantics to constrained the latent variable. To mitigate the non-diverse problem brought by many-to-one, SegCVAE segments multiple prominent semantics to enrich the latent variables. Three novel components, Internal Separation, External Guidance, and Semantic Norms, are proposed to achieve SegCVAE. On dialogue generation tasks, both the automatic and human evaluation results show that SegCVAE achieves new state-of-the-art performance.
Argumentation is used by people both internally, by evaluating arguments and counterarguments to make sense of a situation and take a decision, and externally, e.g., in a debate, by exchanging arguments to reach an agreement or to promote an individual position. In this context, the assessment of the quality of the arguments is of extreme importance, as it strongly influences the evaluation of the overall argumentation, impacting on the decision making process. The automatic assessment of the quality of natural language arguments is recently attracting interest in the Argument Mining field. However, the issue of automatically assessing the quality of an argumentation largely remains a challenging unsolved task. Our contribution is twofold: first, we present a novel resource of 402 student persuasive essays, where three main quality dimensions (i.e., cogency, rhetoric, and reasonableness) have been annotated, leading to 1908 arguments tagged with quality facets; second, we address this novel task of argumentation quality assessment proposing a novel neural architecture based on graph embeddings, that combines both the textual features of the natural language arguments and the overall argument graph, i.e., considering also the support and attack relations holding among the arguments. Results on the persuasive essays dataset outperform state-of-the-art and standard baselines’ performance.
Link prediction is the task of inferring missing links between entities in knowledge graphs. Embedding-based methods have shown effectiveness in addressing this problem by modeling relational patterns in triples. However, the link prediction task often requires contextual information in entity neighborhoods, while most existing embedding-based methods fail to capture it. Additionally, little attention is paid to the diversity of entity representations in different contexts, which often leads to false prediction results. In this situation, we consider that the schema of knowledge graph contains the specific contextual information, and it is beneficial for preserving the consistency of entities across contexts. In this paper, we propose a novel Schema-augmented Multi-level contrastive LEarning framework (SMiLE) to conduct knowledge graph link prediction. Specifically, we first exploit network schema as the prior constraint to sample negatives and pre-train our model by employing a multi-level contrastive learning method to yield both prior schema and contextual information. Then we fine-tune our model under the supervision of individual triples to learn subtler representations for link prediction. Extensive experimental results on four knowledge graph datasets with thorough analysis of each component demonstrate the effectiveness of our proposed framework against state-of-the-art baselines. The implementation of SMiLE is available at https://github.com/GKNL/SMiLE.
Most vision-and-language pretraining research focuses on English tasks. However, the creation of multilingual multimodal evaluation datasets (e.g. Multi30K, xGQA, XVNLI, and MaRVL) poses a new challenge in finding high-quality training data that is both multilingual and multimodal. In this paper, we investigate whether machine translating English multimodal data can be an effective proxy for the lack of readily available multilingual data. We call this framework TD-MML: Translated Data for Multilingual Multimodal Learning, and it can be applied to any multimodal dataset and model. We apply it to both pretraining and fine-tuning data with a state-of-the-art model. In order to prevent models from learning from low-quality translated text, we propose two metrics for automatically removing such translations from the resulting datasets. In experiments on five tasks across 20 languages in the IGLUE benchmark, we show that translated data can provide a useful signal for multilingual multimodal learning, both at pretraining and fine-tuning.
Most of the state-of-the-art methods for abstractive text summarization are under supervised learning settings, while heavily relying on high-quality and large-scale parallel corpora. In this paper, we remove the need for reference summaries and present an unsupervised learning method SCR (Summarize, Contrast and Review) for abstractive summarization, which leverages contrastive learning and is the first work to apply contrastive learning for unsupervised abstractive summarization. Particularly, we use the true source documents as positive source document examples, and strategically generated fake source documents as negative source document examples to train the model to generate good summaries. Furthermore, we consider and improve the writing quality of the generated summaries by guiding them to be similar to human-written texts. The promising results on extensive experiments show that SCR outperforms other unsupervised abstractive summarization baselines, which demonstrates its effectiveness.
Emoji have become a significant part of our informal textual communication. Previous work, addressing the societal and linguistic functions of emoji, overlooked the relation between the semantics and the visual variations of the symbols. In this paper we model and analyze the semantic drift of emoji and discuss the features that may be contributing to the drift, some are unique to emoji and some are more general. Specifically, we explore the relations between graphical changes and semantic changes.
The awareness and mitigation of biases are of fundamental importance for the fair and transparent use of contextual language models, yet they crucially depend on the accurate detection of biases as a precursor. Consequently, numerous bias detection methods have been proposed, which vary in their approach, the considered type of bias, and the data used for evaluation. However, while most detection methods are derived from the word embedding association test for static word embeddings, the reported results are heterogeneous, inconsistent, and ultimately inconclusive. To address this issue, we conduct a rigorous analysis and comparison of bias detection methods for contextual language models. Our results show that minor design and implementation decisions (or errors) have a substantial and often significant impact on the derived bias scores. Overall, we find the state of the field to be both worse than previously acknowledged due to systematic and propagated errors in implementations, yet better than anticipated since divergent results in the literature homogenize after accounting for implementation errors. Based on our findings, we conclude with a discussion of paths towards more robust and consistent bias detection methods.
We propose a multitask pretraining approach ZeroPrompt for zero-shot generalization, focusing on task scaling and zero-shot prompting.While previous models are trained on only a few dozen tasks, we scale to 1,000 tasks for the first time using real-world data. This leads to a crucial discovery that task scaling can be an efficient alternative to model scaling; i.e., the model size has less impact on performance with an extremely large number of tasks. Our results show that task scaling can improve training efficiency by 30 times in FLOPs.Empirically, ZeroPrompt substantially improves both the efficiency and the performance of zero-shot learning across a variety of academic and production datasets.
One of the common traits of past and present approaches for Semantic Role Labeling (SRL) is that they rely upon discrete labels drawn from a predefined linguistic inventory to classify predicate senses and their arguments.However, we argue this need not be the case. In this paper, we present an approach that leverages Definition Modeling to introduce a generalized formulation of SRL as the task of describing predicate-argument structures using natural language definitions instead of discrete labels. Our novel formulation takes a first step towards placing interpretability and flexibility foremost, and yet our experiments and analyses on PropBank-style and FrameNet-style, dependency-based and span-based SRL also demonstrate that a flexible model with an interpretable output does not necessarily come at the expense of performance. We release our software for research purposes at https://github.com/SapienzaNLP/dsrl.
In this paper we conduct a set of experiments aimed to improve our understanding of the lack of semantic isometry in BERT, i.e. the lack of correspondence between the embedding and meaning spaces of its contextualized word representations. Our empirical results show that, contrary to popular belief, the anisotropy is not the root cause of the poor performance of these contextual models’ embeddings in semantic tasks. What does affect both the anisotropy and semantic isometry is a set of known biases: frequency, subword, punctuation, and case. For each one of them, we measure its magnitude and the effect of its removal, showing that these biases contribute but do not completely explain the phenomenon of anisotropy and lack of semantic isometry of these contextual language models.
Multilingual neural machine translation models (MNMT) yield state-of-the-art performance when evaluated on data from a domain and language pair seen at training time. However, when a MNMT model is used to translate under domain shift or to a new language pair, performance drops dramatically. We consider a very challenging scenario: adapting the MNMT model both to a new domain and to a new language pair at the same time. In this paper, we propose m^4Adapter (Multilingual Multi-Domain Adaptation for Machine Translation with a Meta-Adapter), which combines domain and language knowledge using meta-learning with adapters. We present results showing that our approach is a parameter-efficient solution which effectively adapts a model to both a new language pair and a new domain, while outperforming other adapter methods. An ablation study also shows that our approach more effectively transfers domain knowledge across different languages and language information across different domains.
Solving math word problems is the task that analyses the relation of quantities e and requires an accurate understanding of contextual natural language information. Recent studies show that current models rely on shallow heuristics to predict solutions and could be easily misled by small textual perturbations. To address this problem, we propose a Textual Enhanced Contrastive Learning framework, which enforces the models to distinguish semantically similar examples while holding different mathematical logic. We adopt a self-supervised manner strategy to enrich examples with subtle textual variance by textual reordering or problem re-construction. We then retrieve the hardest to differentiate samples from both equation and textual perspectives and guide the model to learn their representations. Experimental results show that our method achieves state-of-the-art on both widely used benchmark datasets and also exquisitely designed challenge datasets in English and Chinese.
Recently, very large pre-trained models achieve state-of-the-art results in various natural language processing (NLP) tasks, but their size makes it more challenging to apply them in resource-constrained environments. Compression techniques allow to drastically reduce the size of the models and therefore their inference time with negligible impact on top-tier metrics. However, the general performance averaged across multiple tasks and/or languages may hide a drastic performance drop on under-represented features, which could result in the amplification of biases encoded by the models. In this work, we assess the impact of compression methods on Multilingual Neural Machine Translation models (MNMT) for various language groups, gender, and semantic biases by extensive analysis of compressed models on different machine translation benchmarks, i.e. FLORES-101, MT-Gender, and DiBiMT. We show that the performance of under-represented languages drops significantly, while the average BLEU metric only slightly decreases. Interestingly, the removal of noisy memorization with compression leads to a significant improvement for some medium-resource languages. Finally, we demonstrate that compression amplifies intrinsic gender and semantic biases, even in high-resource languages.
Building dialogue systems requires a large corpus of annotated dialogues. Such datasets are usually created via crowdsourcing, which is expensive and time-consuming. In this paper, we propose Dialogic, a novel dialogue simulation method based on large language model in-context learning to automate dataset creation. Seeded with a few annotated dialogues, Dialogic automatically selects in-context examples for demonstration and prompts GPT-3 to generate new dialogues and annotations in a controllable way. Our method can rapidly expand a small set of dialogue data with minimum or zero human involvement and parameter update and is thus much more cost-efficient and time-saving than crowdsourcing. Experimental results on the MultiWOZ dataset demonstrate that training a model on the simulated dialogues leads to even better performance than using the same amount of human-generated dialogues under the challenging low-resource settings, with as few as 85 dialogues as a seed. When the full training set is given, our method can still serve as an effective data augmentation method to further improve performance. Human evaluation results also show that our simulated dialogues have near-human fluency and annotation accuracy. The code and data are available at https://github.com/Leezekun/dialogic.
Neural image-to-text radiology report generation systems offer the potential to improve radiology reporting by reducing the repetitive process of report drafting and identifying possible medical errors. These systems have achieved promising performance as measured by widely used NLG metrics such as BLEU and CIDEr. However, the current systems face important limitations. First, they present an increased complexity in architecture that offers only marginal improvements on NLG metrics. Secondly, these systems that achieve high performance on these metrics are not always factually complete or consistent due to both inadequate training and evaluation. Recent studies have shown the systems can be substantially improved by using new methods encouraging 1) the generation of domain entities consistent with the reference and 2) describing these entities in inferentially consistent ways. So far, these methods rely on weakly-supervised approaches (rule-based) and named entity recognition systems that are not specific to the chest X-ray domain. To overcome this limitation, we propose a new method, the RadGraph reward, to further improve the factual completeness and correctness of generated radiology reports. More precisely, we leverage the RadGraph dataset containing annotated chest X-ray reports with entities and relations between entities. On two open radiology report datasets, our system substantially improves the scores up to 14.2% and 25.3% on metrics evaluating the factual correctness and completeness of reports.
A recent line of work in NLP focuses on the (dis)ability of models to generalise compositionally for artificial languages.However, when considering natural language tasks, the data involved is not strictly, or locally, compositional.Quantifying the compositionality of data is a challenging task, which has been investigated primarily for short utterances.We use recursive neural models (Tree-LSTMs) with bottlenecks that limit the transfer of information between nodes.We illustrate that comparing data’s representations in models with and without the bottleneck can be used to produce a compositionality metric.The procedure is applied to the evaluation of arithmetic expressions using synthetic data, and sentiment classification using natural language data.We demonstrate that compression through a bottleneck impacts non-compositional examples disproportionatelyand then use the bottleneck compositionality metric (BCM) to distinguish compositional from non-compositional samples, yielding a compositionality ranking over a dataset.
Timely and effective response to humanitarian crises requires quick and accurate analysis of large amounts of text data – a process that can highly benefit from expert-assisted NLP systems trained on validated and annotated data in the humanitarian response domain. To enable creation of such NLP systems, we introduce and release HumSet, a novel and rich multilingual dataset of humanitarian response documents annotated by experts in the humanitarian response community. The dataset provides documents in three languages (English, French, Spanish) and covers a variety of humanitarian crises from 2018 to 2021 across the globe. For each document, HUMSET provides selected snippets (entries) as well as assigned classes to each entry annotated using common humanitarian information analysis frameworks. HUMSET also provides novel and challenging entry extraction and multi-label entry classification tasks. In this paper, we take a first step towards approaching these tasks and conduct a set of experiments on Pre-trained Language Models (PLM) to establish strong baselines for future research in this domain. The dataset is available at https://blog.thedeep.io/humset/.
Non-autoregressive models achieve significant decoding speedup in neural machine translation but lack the ability to capture sequential dependency. Directed Acyclic Transformer (DA-Transformer) was recently proposed to model sequential dependency with a directed acyclic graph. Consequently, it has to apply a sequential decision process at inference time, which harms the global translation accuracy. In this paper, we present a Viterbi decoding framework for DA-Transformer, which guarantees to find the joint optimal solution for the translation and decoding path under any length constraint. Experimental results demonstrate that our approach consistently improves the performance of DA-Transformer while maintaining a similar decoding speedup.
While fine-tuned language models perform well on many language tasks, they were also shown to rely on superficial surface features such as lexical overlap. Excessive utilization of such heuristics can lead to failure on challenging inputs. We analyze the use of lexical overlap heuristics in natural language inference, paraphrase detection, and reading comprehension (using a novel contrastive dataset),and find that larger models are much less susceptible to adopting lexical overlap heuristics. We also find that longer training leads models to abandon lexical overlap heuristics. Finally, We provide evidence that the disparity between models size has its source in the pre-trained model.
Existing approaches to constructing training data for Natural Language Inference (NLI) tasks, such as for semi-structured table reasoning, are either via crowdsourcing or fully automatic methods. However, the former is expensive and time consuming and thus limits scale, and the latter often produces naive examples that may lack complex reasoning. This paper develops a realistic semi-automated framework for data augmentation for tabular inference. Instead of manually generating a hypothesis for each table, our methodology generates hypothesis templates transferable to similar tables. In addition, our framework entails the creation of rational counterfactual tables based on human written logical constraints and premise paraphrasing. For our case study, we use the INFOTABS (Gupta et al., 2020), which is an entity centric tabular inference dataset. We observed that our framework could generate human-like tabular inference examples, which could benefit training data augmentation, especially in the scenario with limited supervision.
Lexica – words and associated scores – are widely used as simple, interpretable, generalizable language features to predict sentiment, emotions, mental health, and personality. They also provide insight into the psychological features behind those moods and traits. Such lexica, historically created by human experts, are valuable to linguists, psychologists, and social scientists, but they take years of refinement and have limited coverage. In this paper, we investigate how the lexica that provide psycholinguistic insights could be computationally induced and how they should be assessed. We identify generalizability and interpretability as two essential properties of such lexica. We induce lexica using both context-oblivious and context-aware approaches, compare their predictive performance both within the training corpus and across various corpora, and evaluate their quality using crowd-worker assessment. We find that lexica induced from context-oblivious models are more generalizable and interpretable than those from more accurate context-aware transformer models. In addition, lexicon scores can identify explanatory words more reliably than a high performing transformer with feature-importance measures like SHAP.
Transformer language models encode the notion of word order using positional information. Most commonly, this positional information is represented by absolute position embeddings (APEs), that are learned from the pretraining data. However, in natural language, it is not absolute position that matters, but relative position, and the extent to which APEs can capture this type of information has not been studied. In this work, we observe that models trained with APE over-rely on positional information to the point that they break-down when subjected to sentences with shifted position information. Specifically, when models are subjected to sentences starting from a non-zero position (excluding the effect of priming), they exhibit noticeably degraded performance on zero- to full-shot tasks, across a range of model families and model sizes. Our findings raise questions about the efficacy of APEs to model the relativity of position information, and invite further introspection on the sentence and word order processing strategies employed by these models.
Vision-and-dialog navigation is a recent benchmark for evaluating the AI capabilities of perception, interaction, and decision making. While existing methods developed for this benchmark have demonstrated great successes, they mostly rely on large datasets, where data collection can be a challenge, and the learned policies are not adaptive to domain changes. In this paper, we focus on a new problem, referred to as goal-oriented vision-and-dialog navigation (GVDN), where an agent uses reinforcement learning techniques to compute dialog-navigation policies from trial and error. A robot conducts visual navigation to locate target objects, and can talk to a remote human operator as needed. Our remote human is able to provide guidance on navigation only if the robot correctly conveys its location through dialog. Experiments have been conducted using photo-realistic simulation environments. Results suggest that, our agent outperforms competitive baselines in success rate.
Creating challenging tabular inference data is essential for learning complex reasoning. Prior work has mostly relied on two data generation strategies. The first is human annotation, which yields linguistically diverse data but is difficult to scale. The second category for creation is synthetic generation, which is scalable and cost effective but lacks inventiveness. In this research, we present a framework for semi-automatically recasting existing tabular data to make use of the benefits of both approaches. We utilize our framework to build tabular NLI instances from five datasets that were initially intended for tasks like table2text creation, tabular Q/A, and semantic parsing. We demonstrate that recasted data could be used as evaluation benchmarks as well as augmentation data to enhance performance on tabular NLI tasks. Furthermore, we investigate the effectiveness of models trained on recasted data in the zero-shot scenario, and analyse trends in performance across different recasted datasets types.
Large pre-trained language models (PLMs) such as GPT-3 have shown strong in-context learning capabilities, which are highly appealing for domains such as biomedicine that feature high and diverse demands of language technologies but also high data annotation costs. In this paper, we present the first systematic and comprehensive study to compare the few-shot performance of GPT-3 in-context learning with fine-tuning smaller (i.e., BERT-sized) PLMs on two representative biomedical information extraction (IE) tasks: named entity recognition and relation extraction. We follow the true few-shot setting to avoid overestimating models’ few-shot performance by model selection over a large validation set. We also optimize GPT-3’s performance with known techniques such as contextual calibration and dynamic in-context example retrieval. However, our results show that GPT-3 still significantly underperforms compared to simply fine-tuning a smaller PLM. In addition, GPT-3 in-context learning also yields smaller gains in accuracy when more training data becomes available. More in-depth analyses further reveal issues of in-context learning that may be detrimental to IE tasks in general. Given the high cost of experimenting with GPT-3, we hope our study provides helpful guidance for biomedical researchers and practitioners towards more practical solutions such as fine-tuning small PLMs before better in-context learning is available for biomedical IE.
In Transformer-based language models (LMs) the attention mechanism converts token embeddings into contextual embeddings that incorporate information from neighboring words. The resulting contextual hidden state embeddings have enabled highly accurate models of brain responses, suggesting that the attention mechanism constructs contextual embeddings that carry information reflected in language-related brain representations. However, it is unclear whether the attention weights that are used to integrate information across words are themselves related to language representations in the brain. To address this question we analyzed functional magnetic resonance imaging (fMRI) recordings of participants reading English language narratives. We provided the narrative text as input to two LMs (BERT and GPT-2) and extracted their corresponding attention weights. We then used encoding models to determine how well attention weights can predict recorded brain responses. We find that attention weights accurately predict brain responses in much of the frontal and temporal cortices. Our results suggest that the attention mechanism itself carries information that is reflected in brain representations. Moreover, these results indicate cortical areas in which context integration may occur.
Word sense disambiguation (WSD) is the task of identifying the intended sense of a word in context. While prior work on unsupervised WSD has leveraged lexical knowledge bases, such as WordNet and BabelNet, these resources have proven to be less effective for Chinese. Instead, the most widely used lexical knowledge base for Chinese is HowNet. Previous HowNet-based WSD methods have not exploited contextual translation information. In this paper, we present the first HowNet-based WSD system which combines monolingual contextual information from a pretrained neural language model with bilingual information obtained via machine translation and sense translation information from HowNet. The results of our evaluation experiment on a test set from prior work demonstrate that our new method achieves a new state of the art for unsupervised Chinese WSD.
We present Mask-then-Fill, a flexible and effective data augmentation framework for event extraction. Our approach allows for more flexible manipulation of text and thus can generate more diverse data while keeping the original event structure unchanged as much as possible. Specifically, it first randomly masks out an adjunct sentence fragment and then infills a variable-length text span with a fine-tuned infilling model. The main advantage lies in that it can replace a fragment of arbitrary length in the text with another fragment of variable length, compared to the existing methods which can only replace a single word or a fixed-length fragment. On trigger and argument extraction tasks, the proposed framework is more effective than baseline methods and it demonstrates particularly strong results in the low-resource setting. Our further analysis shows that it achieves a good balance between diversity and distributional similarity.
MOBA (Multiplayer Online Battle Arena) games such as Dota2 are currently one of the most popular e-sports gaming genres. Following professional commentaries is a great way to understand and enjoy a MOBA game. However, massive game competitions lack commentaries because of the shortage of professional human commentators. As an alternative, employing machine commentators that can work at any time and place is a feasible solution. Considering the challenges in modeling MOBA games, we propose a data-driven MOBA commentary generation framework, MOBA-E2C, allowing a model to generate commentaries based on the game meta-data. Subsequently, to alleviate the burden of collecting supervised data, we propose a MOBA-FuseGPT generator to generate MOBA game commentaries by fusing the power of a rule-based generator and a generative GPT generator. Finally, in the experiments, we take a popular MOBA game Dota2 as our case and construct a Chinese Dota2 commentary generation dataset Dota2-Commentary. Experimental results demonstrate the superior performance of our approach. To the best of our knowledge, this work is the first Dota2 machine commentator and Dota2-Commentary is the first dataset.
The readability assessment task aims to assign a difficulty grade to a text. While neural models have recently demonstrated impressive performance, most do not exploit the ordinal nature of the difficulty grades, and make little effort for model initialization to facilitate fine-tuning. We address these limitations with soft labels for ordinal regression, and with model pre-training through prediction of pairwise relative text difficulty. We incorporate these two components into a model based on hierarchical attention networks, and evaluate its performance on both English and Chinese datasets. Experimental results show that our proposed model outperforms competitive neural models and statistical classifiers on most datasets.
Recent research on argumentative dialogues has focused on persuading people to take some action, changing their stance on the topic of discussion, or winning debates. In this work, we focus on argumentative dialogues that aim to open up (rather than change) people’s minds to help them become more understanding to views that are unfamiliar or in opposition to their own convictions. To this end, we present a dataset of 183 argumentative dialogues about 3 controversial topics: veganism, Brexit and COVID-19 vaccination. The dialogues were collected using the Wizard of Oz approach, where wizards leverage a knowledge-base of arguments to converse with participants. Open-mindedness is measured before and after engaging in the dialogue using a questionnaire from the psychology literature, and success of the dialogue is measured as the change in the participant’s stance towards those who hold opinions different to theirs. We evaluate two dialogue models: a Wikipedia-based and an argument-based model. We show that while both models perform closely in terms of opening up minds, the argument-based model is significantly better on other dialogue properties such as engagement and clarity.
One reason for the positive impact of Pre-trained Language Models (PLMs) in NLP tasks is their ability to encode semantic types, such as ‘European City’ or ‘Woman’. While previous work has analyzed such information in the context of interpretability, it is not clear how to use types to steer the PLM output. For example, in a cloze statement, it is desirable to steer the model to generate a token that satisfies a user-specified type, e.g., predict a date rather than a location. In this work, we introduce Type Embeddings (TEs), an input embedding that promotes desired types in a PLM. Our proposal is to define a type by a small set of word examples. We empirically study the ability of TEs both in representing types and in steering masking predictions without changes to the prompt text in BERT. Finally, using the LAMA datasets, we show how TEs highly improve the precision in extracting facts from PLMs.
Many word-level adversarial attack approaches for textual data have been proposed in recent studies. However, due to the massive search space consisting of combinations of candidate words, the existing approaches face the problem of preserving the semantics of texts when crafting adversarial counterparts. In this paper, we develop a novel attack strategy to find adversarial texts with high similarity to the original texts while introducing minimal perturbation. The rationale is that we expect the adversarial texts with small perturbation can better preserve the semantic meaning of original texts. Experiments show that, compared with state-of-the-art attack approaches, our approach achieves higher success rates and lower perturbation rates in four benchmark datasets.
Adding interpretability to word embeddings represents an area of active research in textrepresentation. Recent work has explored the potential of embedding words via so-called polardimensions (e.g. good vs. bad, correct vs. wrong). Examples of such recent approachesinclude SemAxis, POLAR, FrameAxis, and BiImp. Although these approaches provide interpretabledimensions for words, they have not been designed to deal with polysemy, i.e. they can not easily distinguish between different senses of words. To address this limitation, we present SensePOLAR, an extension of the original POLAR framework that enables wordsense aware interpretability for pre-trained contextual word embeddings. The resulting interpretable word embeddings achieve a level ofperformance that is comparable to original contextual word embeddings across a variety ofnatural language processing tasks including the GLUE and SQuAD benchmarks. Our workremoves a fundamental limitation of existing approaches by offering users sense aware interpretationsfor contextual word embeddings.
To comprehensibly contextualize decisions, artificial systems in social situations need a high degree of awareness of the rules of conduct of human behavior. Especially transformer-based language models have recently been shown to exhibit some such awareness. But what if norms in some social setting do not adhere to or even blatantly deviate from the mainstream? In this paper, we introduce a novel mechanism based on deontic logic to allow for a flexible adaptation of individual norms by de-biasing training data sets and a task-reduction to textual entailment. Building on the popular ‘Moral Stories’ dataset we on the one hand highlight the intrinsic bias of current language models, on the other hand characterize the adaptability of pre-trained models to deviating norms in fine-tuning settings.
Empathy, which is widely used in psychological counseling, is a key trait of everyday human conversations. Equipped with commonsense knowledge, current approaches to empathetic response generation focus on capturing implicit emotion within dialogue context, where the emotions are treated as a static variable throughout the conversations. However, emotions change dynamically between utterances, which makes previous works difficult to perceive the emotion flow and predict the correct emotion of the target response, leading to inappropriate response. Furthermore, simply importing commonsense knowledge without harmonization may trigger the conflicts between knowledge and emotion, which confuse the model to choose the correct information to guide the generation process. To address the above problems, we propose a Serial Encoding and Emotion-Knowledge interaction (SEEK) method for empathetic dialogue generation. We use a fine-grained encoding strategy which is more sensitive to the emotion dynamics (emotion flow) in the conversations to predict the emotion-intent characteristic of response. Besides, we design a novel framework to model the interaction between knowledge and emotion to solve the conflicts generate more sensible response. Extensive experiments on the utterance-level annotated EMPATHETICDIALOGUES demonstrate that SEEK outperforms the strong baseline in both automatic and manual evaluations.
Knowledge graph (KG) alignment and completion are usually treated as two independent tasks. While recent work has leveraged entity and relation alignments from multiple KGs, such as alignments between multilingual KGs with common entities and relations, a deeper understanding of the ways in which multilingual KG completion (MKGC) can aid the creation of multilingual KG alignments (MKGA) is still limited. Motivated by the observation that structural inconsistencies – the main challenge for MKGA models – can be mitigated through KG completion methods, we propose a novel model for jointly completing and aligning knowledge graphs. The proposed model combines two components that jointly accomplish KG completion and alignment. These two components employ relation-aware graph neural networks that we propose to encode multi-hop neighborhood structures into entity and relation representations. Moreover, we also propose (i) a structural inconsistency reduction mechanism to incorporate information from the completion into the alignment component, and (ii) an alignment seed enlargement and triple transferring mechanism to enlarge alignment seeds and transfer triples during KGs alignment. Extensive experiments on a public multilingual benchmark show that our proposed model outperforms existing competitive baselines, obtaining new state-of-the-art results on both MKGC and MKGA tasks.
This paper describes a framework to automatically generate a spoken question answering (QA) dataset. The framework consists of a question generation (QG) module to generate questions automatically from given text documents, a text-to-speech (TTS) module to convert the text documents into spoken form and an automatic speech recognition (ASR) module to transcribe the spoken content. The final dataset contains question-answer pairs for both the reference text and ASR transcriptions as well as the audio files corresponding to each reference text. For QG and ASR systems we used pre-trained multilingual encoder-decoder transformer models and fine-tuned these models using a limited amount of manually generated QA data and TTS-based speech data, respectively. As a proof of concept, we investigated the proposed framework for Turkish and generated the Turkish Question Answering (TurQuAse) dataset using Wikipedia articles. Manual evaluation of the automatically generated question- answer pairs and QA performance evaluation with state of-the-art models on TurQuAse show that the proposed framework is efficient for automatically generating spoken QA datasets. To the best of our knowledge, TurQuAse is the first publicly available spoken question answering dataset for Turkish. The proposed framework can be easily extended to other languages where a limited amount of QA data is available.
Different from general documents, it is recognised that the ease with which people can understand a biomedical text is eminently varied, owing to the highly technical nature of biomedical documents and the variance of readers’ domain knowledge. However, existing biomedical document summarization systems have paid little attention to readability control, leaving users with summaries that are incompatible with their levels of expertise.In recognition of this urgent demand, we introduce a new task of readability controllable summarization for biomedical documents, which aims to recognise users’ readability demands and generate summaries that better suit their needs: technical summaries for experts and plain language summaries (PLS) for laymen.To establish this task, we construct a corpus consisting of biomedical papers with technical summaries and PLSs written by the authors, and benchmark multiple advanced controllable abstractive and extractive summarization models based on pre-trained language models (PLMs) with prevalent controlling and generation techniques.Moreover, we propose a novel masked language model (MLM) based metric and its variant to effectively evaluate the readability discrepancy between lay and technical summaries.Experimental results from automated and human evaluations show that though current control techniques allow for a certain degree of readability adjustment during generation, the performance of existing controllable summarization methods is far from desirable in this task.
Multimodal fusion addresses the problem of analyzing spoken words in the multimodal context, including visual expressions and prosodic cues. Even when multimodal models lead to performance improvements, it is often unclear whether bimodal and trimodal interactions are learned or whether modalities are processed independently of each other. We propose Multimodal Residual Optimization (MRO) to separate unimodal, bimodal, and trimodal interactions in a multimodal model. This improves interpretability as the multimodal interaction can be quantified. Inspired by Occam’s razor, the main intuition of MRO is that (simpler) unimodal contributions should be learned before learning (more complex) bimodal and trimodal interactions. For example, bimodal predictions should learn to correct the mistakes (residuals) of unimodal predictions, thereby letting the bimodal predictions focus on the remaining bimodal interactions. Empirically, we observe that MRO successfully separates unimodal, bimodal, and trimodal interactions while not degrading predictive performance. We complement our empirical results with a human perception study and observe that MRO learns multimodal interactions that align with human judgments.
For vision-and-language reasoning tasks, both fully connectionist, end-to-end methods and hybrid, neuro-symbolic methods have achieved high in-distribution performance. In which out-of-distribution settings does each paradigm excel? We investigate this question on both single-image and multi-image visual question-answering through four types of generalization tests: a novel segment-combine test for multi-image queries, contrast set, compositional generalization, and cross-benchmark transfer.Vision-and-language end-to-end trained systems exhibit sizeable performance drops across all these tests. Neuro-symbolic methods suffer even more on cross-benchmark transfer from GQA to VQA, but they show smaller accuracy drops on the other generalization tests and their performance quickly improves by few-shot training. Overall, our results demonstrate the complementary benefits of these two paradigms, and emphasize the importance of using a diverse suite of generalization tests to fully characterize model robustness to distribution shift.
Story visualization aims to generate a sequence of images to narrate each sentence in a multi-sentence story, where the images should be realistic and keep global consistency across dynamic scenes and characters. Current works face the problem of semantic misalignment because of their fixed architecture and diversity of input modalities. To address this problem, we explore the semantic alignment between text and image representations by learning to match their semantic levels in the GAN-based generative model. More specifically, we introduce dynamic interactions according to learning to dynamically explore various semantic depths and fuse the different-modal information at a matched semantic level, which thus relieves the text-image semantic misalignment problem. Extensive experiments on different datasets demonstrate the improvements of our approach, neither using segmentation masks nor auxiliary captioning networks, on image quality and story consistency, compared with state-of-the-art methods.
While research on scientific claim verification has led to the development of powerful systems that appear to approach human performance, these approaches have yet to be tested in a realistic setting against large corpora of scientific literature. Moving to this open-domain evaluation setting, however, poses unique challenges; in particular, it is infeasible to exhaustively annotate all evidence documents. In this work, we present SciFact-Open, a new test collection designed to evaluate the performance of scientific claim verification systems on a corpus of 500K research abstracts. Drawing upon pooling techniques from information retrieval, we collect evidence for scientific claims by pooling and annotating the top predictions of four state-of-the-art scientific claim verification models. We find that systems developed on smaller corpora struggle to generalize to SciFact-Open, exhibiting performance drops of at least 15 F1. In addition, analysis of the evidence in SciFact-Open reveals interesting phenomena likely to appear when claim verification systems are deployed in practice, e.g., cases where the evidence supports only a special case of the claim. Our dataset is available at https://github.com/dwadden/scifact-open.
Active learning aims to deliver maximum benefit when resources are scarce. We use COMET-QE, a reference-free evaluation metric, to select sentences for low-resource neural machine translation. Using Swahili, Kinyarwanda and Spanish for our experiments, we show that COMET-QE significantly outperforms two variants of Round Trip Translation Likelihood (RTTL) and random sentence selection by up to 5 BLEU points for 20k sentences selected by Active Learning on a 30k baseline. This suggests that COMET-QE is a powerful tool for sentence selection in the very low-resource limit.
We introduce MedicalSum, a transformer-based sequence-to-sequence architecture for summarizing medical conversations by integrating medical domain knowledge from the Unified Medical Language System (UMLS). The novel knowledge augmentation is performed in three ways: (i) introducing a guidance signal that consists of the medical words in the input sequence, (ii) leveraging semantic type knowledge in UMLS to create clinically meaningful input embeddings, and (iii) making use of a novel weighted loss function that provides a stronger incentive for the model to correctly predict words with a medical meaning. By applying these three strategies, MedicalSum takes clinical knowledge into consideration during the summarization process and achieves state-of-the-art ROUGE score improvements of 0.8-2.1 points (including 6.2% ROUGE-1 error reduction in the PE section) when producing medical summaries of patient-doctor conversations.
The effectiveness of pre-trained language models in downstream tasks is highly dependent on the amount of labeled data available for training. Semi-supervised learning (SSL) is a promising technique that has seen wide attention recently due to its effectiveness in improving deep learning models when training data is scarce. Common approaches employ a teacher-student self-training framework, where a teacher network generates pseudo-labels for unlabeled data, which are then used to iteratively train a student network. In this paper, we propose a new self-training approach for text classification that leverages training dynamics of unlabeled data. We evaluate our approach on a wide range of text classification tasks, including emotion detection, sentiment analysis, question classification and gramaticality, which span a variety of domains, e.g, Reddit, Twitter, and online forums. Notably, our method is successful on all benchmarks, obtaining an average increase in F1 score of 3.5% over strong baselines in low resource settings.
Natural Language Inference (NLI) or Recognizing Textual Entailment (RTE) aims at predicting the relation between a pair of sentences (premise and hypothesis) as entailment, contradiction or semantic independence. Although deep learning models have shown promising performance for NLI in recent years, they rely on large scale expensive human-annotated datasets. Semi-supervised learning (SSL) is a popular technique for reducing the reliance on human annotation by leveraging unlabeled data for training. However, despite its substantial success on single sentence classification tasks where the challenge in making use of unlabeled data is to assign “good enough” pseudo-labels, for NLI tasks, the nature of unlabeled data is more complex: one of the sentences in the pair (usually the hypothesis) along with the class label are missing from the data and require human annotations, which makes SSL for NLI more challenging. In this paper, we propose a novel way to incorporate unlabeled data in SSL for NLI where we use a conditional language model, BART to generate the hypotheses for the unlabeled sentences (used as premises). Our experiments show that our SSL framework successfully exploits unlabeled data and substantially improves the performance of four NLI datasets in low-resource settings. We release our code here: https://github.com/msadat3/SSL_for_NLI
Deidentification seeks to anonymize textual data prior to distribution. Automatic deidentification primarily uses supervised named entity recognition from human-labeled data points. We propose an unsupervised deidentification method that masks words that leak personally-identifying information. The approach utilizes a specially trained reidentification model to identify individuals from redacted personal documents. Motivated by K-anonymity based privacy, we generate redactions that ensure a minimum reidentification rank for the correct profile of the document. To evaluate this approach, we consider the task of deidentifying Wikipedia Biographies, and evaluate using an adversarial reidentification metric. Compared to a set of unsupervised baselines, our approach deidentifies documents more completely while removing fewer words. Qualitatively, we see that the approach eliminates many identifying aspects that would fall outside of the common named entity based approach.
In this work, we combine the two paradigms: Federated Learning (FL) and Continual Learning (CL) for text classification task in cloud-edge continuum. The objective of Federated Continual Learning (FCL) is to improve deep learning models over life time at each client by (relevant and efficient) knowledge transfer without sharing data. Here, we address challenges in minimizing inter-client interference while knowledge sharing due to heterogeneous tasks across clients in FCL setup. In doing so, we propose a novel framework, Federated Selective Inter-client Transfer (FedSeIT) which selectively combines model parameters of foreign clients. To further maximize knowledge transfer, we assess domain overlap and select informative tasks from the sequence of historical tasks at each foreign client while preserving privacy. Evaluating against the baselines, we show improved performance, a gain of (average) 12.4% in text classification over a sequence of tasks using five datasets from diverse domains. To the best of our knowledge, this is the first work that applies FCL to NLP.
In the real world, autonomous driving agents navigate in highly dynamic environments full of unexpected situations where pre-trained models are unreliable. In these situations, what is immediately available to vehicles is often only human operators. Empowering autonomous driving agents with the ability to navigate in a continuous and dynamic environment and to communicate with humans through sensorimotor-grounded dialogue becomes critical. To this end, we introduce Dialogue On the ROad To Handle Irregular Events (DOROTHIE), a novel interactive simulation platform that enables the creation of unexpected situations on the fly to support empirical studies on situated communication with autonomous driving agents. Based on this platform, we created the Situated Dialogue Navigation (SDN), a navigation benchmark of 183 trials with a total of 8415 utterances, around 18.7 hours of control streams, and 2.9 hours of trimmed audio. SDN is developed to evaluate the agent’s ability to predict dialogue moves from humans as well as generate its own dialogue moves and physical navigation actions. We further developed a transformer-based baseline model for these SDN tasks. Our empirical results indicate that language guided-navigation in a highly dynamic environment is an extremely difficult task for end-to-end models. These results will provide insight towards future work on robust autonomous driving agents
In this work, we define a new style transfer task: perspective shift, which reframes a dialouge from informal first person to a formal third person rephrasing of the text. This task requires challenging coreference resolution, emotion attribution, and interpretation of informal text. We explore several baseline approaches and discuss further directions on this task when applied to short dialogues. As a sample application, we demonstrate that applying perspective shifting to a dialogue summarization dataset (SAMSum) substantially improves the zero-shot performance of extractive news summarization models on this data. Additionally, supervised extractive models perform better when trained on perspective shifted data than on the original dialogues. We release our code publicly.
Data augmentation has been a popular method for fine-tuning pre-trained language models to increase model robustness and performance. With augmentation data coming from modifying gold train data (in-sample augmentation) or being harvested from general domain unlabeled data (out-of-sample augmentation), the quality of such data is the key to successful fine-tuning. In this paper, we propose a dynamic data selection method to select effective augmentation data from different augmentation sources according to the model’s learning stage, by identifying a set of augmentation samples that optimally facilitates the learning process of the most current model. The method firstly filters out augmentation samples with noisy pseudo labels through a curriculum learning strategy, then estimates the effectiveness of reserved augmentation data by its influence scores on the current model at every update, allowing the data selection process tightly tailored to model parameters. And the two-stage augmentation strategy considers in-sample augmentation and out-of-sample augmentation in different learning stages. Experiments with both kinds of augmentation data on a variety of sentence classification tasks show that our method outperforms strong baselines, proving the effectiveness of our method. Analysis confirms the dynamic nature of the data effectiveness and the importance of model learning stages in utilization of augmentation data.
Keyphrase generation is the task of generating phrases (keyphrases) that summarize the main topics of a given document. Keyphrases can be either present or absent from the given document. While the extraction of present keyphrases has received much attention in the past, only recently a stronger focus has been placed on the generation of absent keyphrases. However, generating absent keyphrases is challenging; even the best methods show only a modest degree of success. In this paper, we propose a model-agnostic approach called keyphrase dropout (or KPDrop) to improve absent keyphrase generation. In this approach, we randomly drop present keyphrases from the document and turn them into artificial absent keyphrases during training. We test our approach extensively and show that it consistently improves the absent performance of strong baselines in both supervised and resource-constrained semi-supervised settings.
In settings from fact-checking to question answering, we frequently want to know whether a collection of evidence (premises) entails a hypothesis. Existing methods primarily focus on the end-to-end discriminative version of this task, but less work has treated the generative version in which a model searches over the space of statements entailed by the premises to constructively derive the hypothesis. We propose a system for doing this kind of deductive reasoning in natural language by decomposing the task into separate steps coordinated by a search procedure, producing a tree of intermediate conclusions that faithfully reflects the system’s reasoning process. Our experiments on the EntailmentBank dataset (Dalvi et al., 2021) demonstrate that the proposed system can successfully prove true statements while rejecting false ones. Moreover, it produces natural language explanations with a 17% absolute higher step validity than those produced by an end-to-end T5 model.
We propose EnDex, the first human-reaction based model to evaluate dialogue engagingness. EnDex is trained on 80k Reddit-based Engagement Dataset (RED) curated using a novel distant-supervision framework. Engagingness is a key measure that captures high-level quality of AI dialogue systems and closely reflects actual user experience. However, data shortage, plus the abstract and extensive definition of engagingness makes it challenging to develop an automatic metric. Our work departs from mainstream approaches that use synthetic negative examples to train binary classifiers, and instead, proposes a solution using distant-supervision from human-reaction feedback. To support the soundness of our EnDex metric, we offer a theoretical foundation for engagement, an extensive ablation study, and empirical evidence of high correlation on five engagingness related datasets. We will release code, off-the-shelf EnDex model, and a large-scale dataset upon paper publication to facilitate future research.
Weakly supervised text classification methods typically train a deep neural classifier based on pseudo-labels. The quality of pseudo-labels is crucial to final performance but they are inevitably noisy due to their heuristic nature, so selecting the correct ones has a huge potential for performance boost. One straightforward solution is to select samples based on the softmax probability scores in the neural classifier corresponding to their pseudo-labels. However, we show through our experiments that such solutions are ineffective and unstable due to the erroneously high-confidence predictions from poorly calibrated models. Recent studies on the memorization effects of deep neural models suggest that these models first memorize training samples with clean labels and then those with noisy labels. Inspired by this observation, we propose a novel pseudo-label selection method LOPS that takes learning order of samples into consideration. We hypothesize that the learning order reflects the probability of wrong annotation in terms of ranking, and therefore, propose to select the samples that are learnt earlier. LOPS can be viewed as a strong performance-boost plug-in to most existing weakly-supervised text classification methods, as confirmed in extensive experiments on four real-world datasets.
Model compression by way of parameter pruning, quantization, or distillation has recently gained popularity as an approach for reducing the computational requirements of modern deep neural network models for NLP. Inspired by prior works suggesting a connection between simpler, more generalizable models and those that lie within wider loss basins, we hypothesize that optimizing for flat minima should lead to simpler parameterizations and thus more compressible models. We propose to combine sharpness-aware minimization (SAM) with various task-specific model compression methods, including iterative magnitude pruning (IMP), structured pruning with a distillation objective, and post-training dynamic quantization. Empirically, we show that optimizing for flatter minima consistently leads to greater compressibility of parameters compared to vanilla Adam when fine-tuning BERT models, with little to no loss in accuracy on the GLUE text classification and SQuAD question answering benchmarks. Moreover, SAM finds superior winning tickets during IMP that 1) are amenable to vanilla Adam optimization, and 2) transfer more effectively across tasks.
Zero-shot multi-label text classification (ZMTC) is a fundamental task in natural language processing with applications in the cold start problem of recommendation systems. Ideally, one would learn an expressive representation of both input text and label features so that ZMTC is transformed into a nearest neighbor search problem. However, the existing representation learning approaches for ZMTC struggle with accuracy as well as poor training efficiency. Firstly, the input text is structural, consisting of both short title sentences and long content paragraphs. It is challenging to model the correlation between short label descriptions and long structural input documents. Secondly, the enormous label space in ZMTC forces the existing approaches to perform multi-stage learning with label engineering. As a result, the training overhead is significant. In this paper, we address both problems by introducing an end-to-end structural contrastive representation learning approach. We propose a randomized text segmentation (RTS) technique to generate high-quality contrastive pairs. This RTS technique allows us to model title-content correlation. Additionally, we simplify the multi-stage ZMTC learning strategy by avoiding label engineering. Extensive experiments demonstrate that our approach leads to up to 2.33% improvement in precision@1 and 5.94x speedup in training time on publicly available datasets. Our code is available publicly.
Knowledge Distillation (KD) is a commonly used technique for improving the generalization of compact Pre-trained Language Models (PLMs) on downstream tasks. However, such methods impose the additional burden of training a separate teacher model for every new dataset.Alternatively, one may directly work on the improvement of the optimization procedure of the compact model towards better generalization. Recent works observe that the flatness of the local minimum correlates well with better generalization.In this work, we adapt Stochastic Weight Averaging (SWA), a method encouraging convergence to a flatter minimum, to fine-tuning PLMs. We conduct extensive experiments on various NLP tasks (text classification, question answering, and generation) and different model architectures and demonstrate that our adaptation improves the generalization without extra computation cost. Moreover, we observe that this simple optimization technique is able to outperform the state-of-the-art KD methods for compact models.
Language models pre-trained on large self-supervised corpora, followed by task-specific fine-tuning has become the dominant paradigm in NLP. These pre-training datasets often have a one-to-many structure—e.g. in dialogue there are many valid responses for a given context. However, only some of these responses will be desirable in our downstream task. This raises the question of how we should train the model such that it can emulate the desirable behaviours, but not the undesirable ones. Current approaches train in a one-to-one setup—only a single target response is given for a single dialogue context—leading to models only learning to predict the average response, while ignoring the full range of possible responses. Using text-based games as a testbed, our approach, PASA, uses discrete latent variables to capture the range of different behaviours represented in our larger pre-training dataset. We then use knowledge distillation to distil the posterior probability distribution into a student model. This probability distribution is far richer than learning from only the hard targets of the dataset, and thus allows the student model to benefit from the richer range of actions the teacher model has learned. Results show up to 49% empirical improvement over the previous state-of-the-art model on the Jericho Walkthroughs dataset.
A growing body of research has demonstrated the inability of NLP models to generalize compositionally and has tried to alleviate it through specialized architectures, training schemes, and data augmentation, among other approaches. In this work, we study a different approach: training on instances with diverse structures. We propose a model-agnostic algorithm for subsampling such sets of instances from a labeled instance pool with structured outputs. Evaluating on both compositional template splits and traditional IID splits of 5 semantic parsing datasets of varying complexity, we show that structurally diverse training using our algorithm leads to comparable or better generalization than prior algorithms in 9 out of 10 dataset-split type pairs. In general, we find structural diversity to consistently improve sample efficiency compared to random train sets. Moreover, we show that structurally diverse sampling yields comprehensive test sets that are a lot more challenging than IID test sets. Finally, we provide two explanations for improved generalization from diverse train sets: 1) improved coverage of output substructures, and 2) a reduction in spurious correlations between these substructures.
Text summarization is a user-preference based task, i.e., for one document, users often have different priorities for the summary. As a key aspect of customization in summarization, granularity is used to measure the semantic coverage between the summary and source document. However, developing systems that can generate summaries with customizable semantic coverage is still an under-explored topic. In this paper, we propose the first unsupervised multi-granularity summarization framework, GranuSum. We take events as the basic semantic units of the source documents and propose to rank these events by their salience. We also develop a model to summarize input documents with given events as anchors and hints. By inputting different numbers of events, GranuSum is capable of producing multi-granular summaries in an unsupervised manner. Meanwhile, we annotate a new benchmark GranuDUC that contains multiple summaries at different granularities for each document cluster. Experimental results confirm the substantial superiority of GranuSum on multi-granularity summarization over strong baselines. Furthermore, by exploiting the event information, GranuSum also exhibits state-of-the-art performance under the conventional unsupervised abstractive setting.
We propose Heuristic Guided Lookahead Decoding (HeLo), a novel decoding strategy for conversation infilling. Conversation infilling aims to generate a seamless bridge of utterances connecting a given pair of source and target utterances. HeLo does not require fine-tuning or extra models – only the generating model itself. Instead, HeLo leverages a greedy lookahead phase before committing to any token. The HeLo framework is simple and can augment conventional decoding strategies paired with any autoregressive language model. Smooth transitions between utterances are encouraged with an annealing schedule. Our experiments show HeLo outperforms several baselines when evaluated with both automatic and human evaluation metrics, which, we argue, are appropriate for the task.
Word embedding aims to learn the dense representation of words and has become a regular input preparation in many NLP tasks. Due to the data and computation intensive nature of learning embeddings from scratch, a more affordable way is to borrow the pretrained embedding available in public and fine-tune the embedding through a domain specific downstream dataset. A privacy concern can arise if a malicious owner of the pretrained embedding gets access to the fine-tuned embedding and tries to infer the critical information from the downstream datasets. In this study, we propose a novel embedding inversion framework called Invernet that materializes the privacy concern by inferring the context distribution in the downstream dataset, which can lead to key information breach. With extensive experimental studies on two real-world news datasets: Antonio Gulli’s News and New York Times, we validate the feasibility of proposed privacy attack and demonstrate the effectiveness of Invernet on inferring downstream datasets based on multiple word embedding methods.
Natural language inference has trended toward studying contexts beyond the sentence level. An important application area is law: past cases often do not foretell how they apply to new situations and implications must be inferred. This paper introduces LawngNLI, constructed from U.S. legal opinions with automatic labels with high human-validated accuracy. Premises are long and multigranular. Experiments show two use cases. First, LawngNLI can benchmark for in-domain generalization from short to long contexts. It has remained unclear if large-scale long-premise NLI datasets actually need to be constructed: near-top performance on long premises could be achievable by fine-tuning using short premises. Without multigranularity, benchmarks cannot distinguish lack of fine-tuning on long premises versus domain shift between short and long datasets. In contrast, our long and short premises share the same examples and domain. Models fine-tuned using several past NLI datasets and/or our short premises fall short of top performance on our long premises. So for at least certain domains (such as ours), large-scale long-premise datasets are needed. Second, LawngNLI can benchmark for implication-based retrieval. Queries are entailed or contradicted by target documents, allowing users to move between arguments and evidence. Leading retrieval models perform reasonably zero shot on a LawngNLI-derived retrieval task. We compare different systems for re-ranking, including lexical overlap and cross-encoders fine-tuned using a modified LawngNLI or past NLI datasets. LawngNLI can train and test systems for implication-based case retrieval and argumentation.
How can we protect the intellectual property of trained NLP models? Modern NLP models are prone to stealing by querying and distilling from their publicly exposed APIs. However, existing protection methods such as watermarking only work for images but are not applicable to text. We propose Distillation-Resistant Watermarking (DRW), a novel technique to protect NLP models from being stolen via distillation. DRW protects a model by injecting watermarks into the victim’s prediction probability corresponding to a secret key and is able to detect such a key by probing a suspect model. We prove that a protected model still retains the original accuracy within a certain bound. We evaluate DRW on a diverse set of NLP tasks including text classification, part-of-speech tagging, and named entity recognition. Experiments show that DRW protects the original model and detects stealing suspects at 100% mean average precision for all four tasks while the prior method fails on two.
While counterfactual data augmentation offers a promising step towards robust generalization in natural language processing, producing a set of counterfactuals that offer valuable inductive bias for models remains a challenge. Most existing approaches for producing counterfactuals, manual or automated, rely on small perturbations via minimal edits, resulting in simplistic changes. We introduce NeuroCounterfactuals, designed as loose counterfactuals, allowing for larger edits which result in naturalistic generations containing linguistic diversity, while still bearing similarity to the original document. Our novel generative approach bridges the benefits of constrained decoding, with those of language model adaptation for sentiment steering. Training data augmentation with our generations results in both in-domain and out-of-domain improvements for sentiment classification, outperforming even manually curated counterfactuals, under select settings. We further present detailed analyses to show the advantages of NeuroCounterfactuals over approaches involving simple, minimal edits.
Transformer-based pre-trained models are known to encode societal biases not only in their contextual representations, but also in downstream predictions when fine-tuned on task-specific data.We present D-Bias, an approach that selectively eliminates stereotypical associations (e.g, co-occurrence statistics) at fine-tuning, such that the model doesn’t learn to excessively rely on those signals.D-Bias attenuates biases from both identity words and frequently co-occurring proxies, which we select using pointwise mutual information.We apply D-Bias to a) occupation classification, and b) toxicity classification and find that our approach substantially reduces downstream biases (e.g. by > 60% in toxicity classification, for identities that are most frequently flagged as toxic on online platforms).In addition, we show that D-Bias dramatically improves upon scrubbing, i.e., removing only the identity words in question.We also demonstrate that D-Bias easily extends to multiple identities, and achieves competitive performance with two recently proposed debiasing approaches: R-LACE and INLP.
Numerous works use word embedding-based metrics to quantify societal biases and stereotypes in texts. Recent studies have found that word embeddings can capture semantic similarity but may be affected by word frequency. In this work we study the effect of frequency when measuring female vs. male gender bias with word embedding-based bias quantification methods. We find that Skip-gram with negative sampling and GloVe tend to detect male bias in high frequency words, while GloVe tends to return female bias in low frequency words. We show these behaviors still exist when words are randomly shuffled. This proves that the frequency-based effect observed in unshuffled corpora stems from properties of the metric rather than from word associations. The effect is spurious and problematic since bias metrics should depend exclusively on word co-occurrences and not individual word frequencies. Finally, we compare these results with the ones obtained with an alternative metric based on Pointwise Mutual Information. We find that this metric does not show a clear dependence on frequency, even though it is slightly skewed towards male bias across all frequencies.
Natural language inference (NLI) is critical in many domains requiring complex decision-making, such as the biomedical domain. We introduce a novel semi-supervised procedure that bootstraps biomedical NLI datasets from positive entailment examples present in abstracts of biomedical publications. We focus on challenging texts where the hypothesis includes mechanistic information such as biochemical interactions between two entities. A key contribution of this work is automating the creation of negative examples that are informative without being simplistic. We generate a range of negative examples using nine strategies that manipulate the structure of the underlying mechanisms both with rules, e.g., flip the roles of the entities in the interaction, and, more importantly, by imposing the perturbed conditions as logical constraints in a neuro-logical decoding system (CITATION).We use this procedure to create a novel dataset for NLI in the biomedical domain, called . The accuracy of neural classifiers on this dataset is in the mid 70s F1, which indicates that this NLI task remains to be solved. Critically, we observe that the performance on the different classes of negative examples varies widely, from 97% F1 on the simple negative examples that change the role of the entities in the hypothesis, to barely better than chance on the negative examples generated using neuro-logic decoding.
Multimodal speech emotion recognition (SER) and sentiment analysis (SA) are important techniques for human-computer interaction. Most existing multimodal approaches utilize either shallow cross-modal fusion of pretrained features, or deep cross-modal fusion with raw features. Recently, attempts have been made to fuse pretrained feature representations in a deep fusion manner during fine-tuning stage. However those approaches have not led to improved results, partially due to their relatively simple fusion mechanisms and lack of proper cross-modal pretraining. In this work, leveraging single-modal pretrained models (RoBERTa and HuBERT), we propose a novel deeply-fused audio-text bi-modal transformer with carefully designed cross-modal fusion mechanism and a stage-wise cross-modal pretraining scheme to fully facilitate the cross-modal learning. Our experiment results show that the proposed method achieves state-of-the-art results on the public IEMOCAP emotion and CMU-MOSEI sentiment datasets, exceeding the previous benchmarks by a large margin.
Conversations on social media tend to go off-topic and turn into different and sometimes toxic exchanges. Previous work focuses on analysing textual dialogues that have derailed into toxic content, but the range of derailment types is much broader, including spam or bot content, tangential comments, etc. In addition, existing work disregards conversations that involve visual information (i.e. images or videos), which are prevalent on most platforms. In this paper, we take a broader view of conversation derailment and propose a new challenge: detecting derailment based on the “change of conversation topic”, where the topic is defined by an initial post containing both a text and an image. For that, we (i) create the first Multimodal Conversation Derailment (MCD) dataset, and (ii) introduce a new multimodal conversational architecture (MMConv) that utilises visual and conversational contexts to classify comments for derailment. Experiments show that MMConv substantially outperforms previous text-based approaches to detect conversation derailment, as well as general multimodal classifiers. MMConv is also more robust to textual noise, since it relies on richer contextual information.
Construction of human-curated annotated datasets for abstractive text summarization (ATS) is very time-consuming and expensive because creating each instance requires a human annotator to read a long document and compose a shorter summary that would preserve the key information relayed by the original document. Active Learning (AL) is a technique developed to reduce the amount of annotation required to achieve a certain level of machine learning model performance. In information extraction and text classification, AL can reduce the amount of labor up to multiple times. Despite its potential for aiding expensive annotation, as far as we know, there were no effective AL query strategies for ATS. This stems from the fact that many AL strategies rely on uncertainty estimation, while as we show in our work, uncertain instances are usually noisy, and selecting them can degrade the model performance compared to passive annotation. We address this problem by proposing the first effective query strategy for AL in ATS based on diversity principles. We show that given a certain annotation budget, using our strategy in AL annotation helps to improve the model performance in terms of ROUGE and consistency scores. Additionally, we analyze the effect of self-learning and show that it can additionally increase the performance of the model.
Memorization presents a challenge for several constrained Natural Language Generation (NLG) tasks such as Neural Machine Translation (NMT), wherein the proclivity of neural models to memorize noisy and atypical samples reacts adversely with the noisy (web crawled) datasets. However, previous studies of memorization in constrained NLG tasks have only focused on counterfactual memorization, linking it to the problem of hallucinations. In this work, we propose a new, inexpensive algorithm for extractive memorization (exact training data generation under insufficient context) in constrained sequence generation tasks and use it to study extractive memorization and its effects in NMT. We demonstrate that extractive memorization poses a serious threat to NMT reliability by qualitatively and quantitatively characterizing the memorized samples as well as the model behavior in their vicinity. Based on empirical observations, we develop a simple algorithm which elicits non-memorized translations of memorized samples from the same model, for a large fraction of such samples. Finally, we show that the proposed algorithm could also be leveraged to mitigate memorization in the model through finetuning. We have released the code to reproduce our results at https://github.com/vyraun/Finding-Memo.
Traditional machine translation (MT) metrics provide an average measure of translation quality that is insensitive to the long tail of behavioral problems. Examples include translation of numbers, physical units, dropped content and hallucinations. These errors, which occur rarely and unpredictably in Neural Machine Translation (NMT), greatly undermine the reliability of state-of-the-art MT systems. Consequently, it is important to have visibility into these problems during model development.Towards this end, we introduce SALTED, a specifications-based framework for behavioral testing of NMT models. At the core of our approach is the use of high-precision detectors that flag errors (or alternatively, verify output correctness) between a source sentence and a system output. These detectors provide fine-grained measurements of long-tail errors, providing a trustworthy view of problems that were previously invisible. We demonstrate that such detectors could be used not just to identify salient long-tail errors in MT systems, but also for higher-recall filtering of the training data, fixing targeted errors with model fine-tuning in NMT and generating novel data for metamorphic testing to elicit further bugs in models.
There are many potential benefits to news readers accessing diverse sources. Modern news aggregators do the hard work of organizing the news, offering readers a plethora of source options, but choosing which source to read remains challenging.We propose a new framework to assist readers in identifying source differences and gaining an understanding of news coverage diversity.The framework is based on the generation of Discord Questions: questions with a diverse answer pool, explicitly illustrating source differences.To assemble a prototype of the framework, we focus on two components: (1) discord question generation, the task of generating questions answered differently by sources, for which we propose an automatic scoring method, and create a model that improves performance from current question generation (QG) methods by 5%, (2) answer consolidation, the task of grouping answers to a question that are semantically similar, for which we collect data and repurpose a method that achieves 81% balanced accuracy on our realistic test set.We illustrate the framework’s feasibility through a prototype interface. Even though model performance at discord QG still lags human performance by more than 15%, generated questions are judged to be more interesting than factoid questions and can reveal differences in the level of detail, sentiment, and reasoning of sources in news coverage. Code is available at https://github.com/Salesforce/discord_questions.
We introduce question answering with a cotext in focus, a task that simulates a free interaction with a QA system. The user reads on a screen some information about a topic, and they can follow-up with questions that can be either related or not to the topic; and the answer can be found in the document containing the screen content or from other pages. We call such information context. To study the task, we construct FocusQA, a dataset for answer sentence selection (AS2) with 12,165011unique question/context pairs, and a total of 109,940 answers. To build the dataset, we developed a novel methodology that takes existing questions and pairs them with relevant contexts. To show the benefits of this approach, we present a comparative analysis with a set of questions written by humans after reading the context, showing that our approach greatly helps in eliciting more realistic question/context pairs. Finally, we show that the task poses several challenges for incorporating contextual information. In this respect, we introduce strong baselines for answer sentence selection that outperform the precision of state-of-the-art models for AS2 up to 21.3% absolute points.
NLP research on public opinion manipulation campaigns has primarily focused on detecting overt strategies such as fake news and disinformation. However, information manipulation in the ongoing Russia-Ukraine war exemplifies how governments and media also employ more nuanced strategies. We release a new dataset, VoynaSlov, containing 38M+ posts from Russian media outlets on Twitter and VKontakte, as well as public activity and responses, immediately preceding and during the 2022 Russia-Ukraine war. We apply standard and recently-developed NLP models on VoynaSlov to examine agenda setting, framing, and priming, several strategies underlying information manipulation, and reveal variation across media outlet control, social media platform, and time. Our examination of these media effects and extensive discussion of current approaches’ limitations encourage further development of NLP models for understanding information manipulation in emerging crises, as well as other real-world and interdisciplinary tasks.
Few-Shot Text Classification (FSTC) imitates humans to learn a new text classifier efficiently with only few examples, by leveraging prior knowledge from historical tasks. However, most prior works assume that all the tasks are sampled from a single data source, which cannot adapt to real-world scenarios where tasks are heterogeneous and lie in different distributions. As such, existing methods may suffer from their globally knowledge-shared mechanisms to handle the task heterogeneity. On the other hand, inherent task relationships are not explicitly captured, making task knowledge unorganized and hard to transfer to new tasks. Thus, we explore a new FSTC setting where tasks can come from a diverse range of data sources. To address the task heterogeneity, we propose a self-supervised hierarchical task clustering (SS-HTC) method. SS-HTC not only customizes the cluster-specific knowledge by dynamically organizing heterogeneous tasks into different clusters in hierarchical levels but also disentangles the underlying relations between tasks to improve the interpretability. Empirically, extensive experiments on five public FSTC benchmark datasets demonstrate the effectiveness of SS-HTC.
In-context learning using large language models has recently shown surprising results for semantic parsing tasks such as Text-to-SQL translation.Prompting GPT-3 or Codex using several examples of question-SQL pairs can produce excellent results, comparable to state-of-the-art finetuning-based models.However, existing work primarily focuses on English datasets, and it is unknown whether large language models can serve as competitive semantic parsers for other languages.To bridge this gap, our work focuses on cross-lingual Text-to-SQL semantic parsing for translating non-English utterances into SQL queries based on an English schema.We consider a zero-shot transfer learning setting with the assumption that we do not have any labeled examples in the target language (but have annotated examples in English).This work introduces the XRICL framework, which learns to retrieve relevant English exemplars for a given query to construct prompts.We also include global translation exemplars for a target language to facilitate the translation process for large language models.To systematically evaluate our model, we construct two new benchmark datasets, XSpider and XKaggle-dbqa, which include questions in Chinese, Vietnamese, Farsi, and Hindi.Our experiments show that XRICL effectively leverages large pre-trained language models to outperform existing baselines.Data and code are publicly available at https://github.com/Impavidity/XRICL.
Knowledge Distillation (KD) has been extensively used for natural language understanding (NLU) tasks to improve a small model’s (a student) generalization by transferring the knowledge from a larger model (a teacher). Although KD methods achieve state-of-the-art performance in numerous settings, they suffer from several problems limiting their performance. It is shown in the literature that the capacity gap between the teacher and the student networks can make KD ineffective. Additionally, existing KD techniques do not mitigate the noise in the teacher’s output: modeling the noisy behaviour of the teacher can distract the student from learning more useful features. We propose a new KD method that addresses these problems and facilitates the training compared to previous techniques. Inspired by continuation optimization, we design a training procedure that optimizes the highly non-convex KD objective by starting with the smoothed version of this objective and making it more complex as the training proceeds. Our method (Continuation-KD) achieves state-of-the-art performance across various compact architectures on NLU (GLUE benchmark) and computer vision tasks (CIFAR-10 and CIFAR-100).
Neuropsychological exams are commonly used to diagnose various kinds of cognitive impairment. They typically involve a trained examiner who conducts a series of cognitive tests with a subject. In recent years, there has been growing interest in developing machine learning methods to extract speech and language biomarkers from exam recordings to provide automated input for cognitive assessment. Inspired by recent findings suggesting that the examiner’s language can influence cognitive impairment classifications, in this paper, we study the influence of the examiner on automatic dementia identification decisions in real-world neuropsychological exams. To mitigate the influence of the examiner, we propose a systematic three-stage pipeline for detecting dementia from exam recordings. In the first stage, we perform audio-based speaker diarization (i.e., estimating who spoke when?) by incorporating speaker discriminative features. In the second stage, we employ text-based language models to identify the role of the speaker (i.e., examiner or subject). Finally, in the third stage, we employ text- and audio-based models to detect cognitive impairment from hypothesized subject segments. Our studies suggest that incorporating audio-based diarization followed by text-based role identification helps mitigate the influences from the examiner’s segments. Further, we found that the text and audio modalities complement each other, and the performance improves when we use both modalities. We also perform several carefully designed experimental studies to assess the performance of each stage.
Recently, author-annotated sarcasm datasets, which focus on intended, rather than perceived sarcasm, have been introduced. Although datasets collected using first-party annotation have important benefits, there is no comparison of human and machine performance on these new datasets. In this paper, we collect new annotations to provide human-level benchmarks for these first-party annotated sarcasm tasks in both English and Arabic, and compare the performance of human annotators to that of state-of-the-art sarcasm detection systems. Our analysis confirms that sarcasm detection is extremely challenging, with individual humans performing close to or slightly worse than the best trained models. With majority voting, however, humans are able to achieve the best results on all tasks. We also perform error analysis, finding that some of the most challenging examples are those that require additional context. We also highlight common features and patterns used to express sarcasm in English and Arabic such as idioms and proverbs. We suggest that to better capture sarcasm, future sarcasm detection datasets and models should focus on representing conversational and cultural context while leveraging world knowledge and common sense.
We focus on the cross-lingual Text-to-SQL semantic parsing task,where the parsers are expected to generate SQL for non-English utterances based on English database schemas.Intuitively, English translation as side information is an effective way to bridge the language gap,but noise introduced by the translation system may affect parser effectiveness.In this work, we propose a Representation Mixup Framework (Rex) for effectively exploiting translations in the cross-lingual Text-to-SQL task.Particularly, it uses a general encoding layer, a transition layer, and a target-centric layer to properly guide the information flow of the English translation.Experimental results on CSpider and VSpider show that our framework can benefit from cross-lingual training and improve the effectiveness of semantic parsers, achieving state-of-the-art performance.
JamPatoisNLI provides the first dataset for natural language inference in a creole language, Jamaican Patois.Many of the most-spoken low-resource languages are creoles. These languages commonly have a lexicon derived from a major world language and a distinctive grammar reflecting the languages of the original speakers and the process of language birth by creolization. This gives them a distinctive place in exploring the effectiveness of transfer from large monolingual or multilingual pretrained models. While our work, along with previous work, shows that transfer from these models to low-resource languages that are unrelated to languages in their training set is not very effective, we would expect stronger results from transfer to creoles. Indeed, our experiments show considerably better results from few-shot learning of JamPatoisNLI than for such unrelated languages, and help us begin to understand how the unique relationship between creoles and their high-resource base languages affect cross-lingual transfer. JamPatoisNLI, which consists of naturally-occurring premises and expert-written hypotheses, is a step towards steering research into a traditionally underserved language and a useful benchmark for understanding cross-lingual NLP.
Recently, the relationship between automated and human evaluation of topic models has been called into question. Method developers have staked the efficacy of new topic model variants on automated measures, and their failure to approximate human preferences places these models on uncertain ground. Moreover, existing evaluation paradigms are often divorced from real-world use.Motivated by content analysis as a dominant real-world use case for topic modeling, we analyze two related aspects of topic models that affect their effectiveness and trustworthiness in practice for that purpose: the stability of their estimates and the extent to which the model’s discovered categories align with human-determined categories in the data. We find that neural topic models fare worse in both respects compared to an established classical method. We take a step toward addressing both issues in tandem by demonstrating that a straightforward ensembling method can reliably outperform the members of the ensemble.
Recent works that revealed the vulnerability of dialogue state tracking (DST) models to distributional shifts have made holistic comparisons on robustness and qualitative analyses increasingly important for understanding their relative performance. We present our findings from standardized and comprehensive DST diagnoses, which have previously been sparse and uncoordinated, using our toolkit, CheckDST, a collection of robustness tests and failure mode analytics. We discover that different classes of DST models have clear strengths and weaknesses, where generation models are more promising for handling language variety while span-based classification models are more robust to unseen entities. Prompted by this discovery, we also compare checkpoints from the same model and find that the standard practice of selecting checkpoints using validation loss/accuracy is prone to overfitting and each model class has distinct patterns of failure. Lastly, we demonstrate how our diagnoses motivate a pre-finetuning procedure with non-dialogue data that offers comprehensive improvements to generation models by alleviating the impact of distributional shifts through transfer learning.
We propose a novel open-domain question answering (ODQA) framework for answering single/multi-hop questions across heterogeneous knowledge sources.The key novelty of our method is the introduction of the intermediary modules into the current retriever-reader pipeline.Unlike previous methods that solely rely on the retriever for gathering all evidence in isolation,our intermediary performs a chain of reasoning over the retrieved set.Specifically, our method links the retrieved evidence with its related global context into graphs and organizes them into a candidate list of evidence chains.Built upon pretrained language models, our system achieves competitive performance on two ODQA datasets, OTT-QA and NQ, against tables and passages from Wikipedia.In particular, our model substantially outperforms the previous state-of-the-art on OTT-QA with an exact match score of 47.3 (45% relative gain).
Providing better language tools for low-resource and endangered languages is imperative for equitable growth.Recent progress with massively multilingual pretrained models has proven surprisingly effective at performing zero-shot transfer to a wide variety of languages.However, this transfer is not universal, with many languages not currently understood by multilingual approaches.It is estimated that only 72 languages possess a “small set of labeled datasets” on which we could test a model’s performance, the vast majority of languages not having the resources available to simply evaluate performances on.In this work, we attempt to clarify which languages do and do not currently benefit from such transfer.To that end, we develop a general approach that requires only unlabelled text to detect which languages are not well understood by a cross-lingual model.Our approach is derived from the hypothesis that if a model’s understanding is insensitive to perturbations to text in a language, it is likely to have a limited understanding of that language.We construct a cross-lingual sentence similarity task to evaluate our approach empirically on 350, primarily low-resource, languages.
Humor is an inherently social phenomenon, with humorous utterances shaped by what is socially and culturally accepted. Understanding humor is an important NLP challenge, with many applications to human-computer interactions. In this work we explore humor in the context of Cards Against Humanity – a party game where players complete fill-in-the-blank statements using cards that can be offensive or politically incorrect.We introduce a novel dataset of 300,000 online games of Cards Against Humanity, including 785K unique jokes, analyze it and provide insights. We trained machine learning models to predict the winning joke per game, achieving performance twice as good (20%) as random, even without any user information.On the more difficult task of judging novel cards, we see the models’ ability to generalize is moderate. Interestingly, we find that our models are primarily focused on punchline card, with the context having little impact.Analyzing feature importance, we observe that short, crude, juvenile punchlines tend to win.
The argument role in event extraction refers to the relation between an event and an argument participating in it. Despite the great progress in event extraction, existing studies still depend on roles pre-defined by domain experts. These studies expose obvious weakness when extending to emerging event types or new domains without available roles. Therefore, more attention and effort needs to be devoted to automatically customizing argument roles. In this paper, we define this essential but under-explored task: open-vocabulary argument role prediction. The goal of this task is to infer a set of argument roles for a given event type. We propose a novel unsupervised framework, RolePred for this task. Specifically, we formulate the role prediction problem as an in-filling task and construct prompts for a pre-trained language model to generate candidate roles. By extracting and analyzing the candidate arguments, the event-specific roles are further merged and selected. To standardize the research of this task, we collect a new human-annotated event extraction dataset including 143 customized argument roles with rich semantics. On this dataset, RolePred outperforms the existing methods by a large margin.
End-to-end spoken language understanding (SLU) systems are gaining popularity over cascaded approaches due to their simplicity and ability to avoid error propagation. However, these systems model sequence labeling as a sequence prediction task causing a divergence from its well-established token-level tagging formulation. We build compositional end-to-end SLU systems that explicitly separate the added complexity of recognizing spoken mentions in SLU from the NLU task of sequence labeling. By relying on intermediate decoders trained for ASR, our end-to-end systems transform the input modality from speech to token-level representations that can be used in the traditional sequence labeling framework. This composition of ASR and NLU formulations in our end-to-end SLU system offers direct compatibility with pre-trained ASR and NLU systems, allows performance monitoring of individual components and enables the use of globally normalized losses like CRF, making them attractive in practical scenarios. Our models outperform both cascaded and direct end-to-end models on a labeling task of named entity recognition across SLU benchmarks.
Neural language models have been analyzed for their linguistic and extra-linguistic knowledge via probing. Of particular interest has been the following question: how much can a language model trained only on form learn about meaning? Recent work has demonstrated via probing classifiers that in the setting of simple procedural text, where by “meaning” we mean the underlying world state, language models have a non-trivial performance on world state tracking. However, our proposed evaluation based on model predictions shows differing results, suggesting that these models are either not capturing the world state or not using it. How do these results change if the model has access to the world state? We explore this alternate setting with access to the underlying world state only during training and investigate ways of “baking in” the state knowledge along with the primary task of language modeling. Our proposed approaches allow for state probing during inference simply via text prompts, avoiding any probing classifier machinery. In terms of performance, we show that baking in the state knowledge during training leads to significant improvements in state tracking performance and text generation quality,
In the past few years, large pre-trained language models (PLMs) have been widely adopted in different areas and have made fundamental improvements over a variety of downstream tasks in natural language processing (NLP). Meanwhile, domain-specific variants of PLMs are being proposed to address the needs of domains that demonstrate a specific pattern of writing and vocabulary, e.g., BioBERT for the biomedical domain and ClinicalBERT for the clinical domain. Recently, generative language models like BART and T5 are gaining popularity with their competitive performance on text generation as well as on tasks cast as generative problems. However, in the clinical domain, such domain-specific generative variants are still underexplored. To address this need, our work introduces a T5-based text-to-text transformer model pre-trained on clinical text, i.e., ClinicalT5. We evaluate the proposed model both intrinsically and extrinsically over a diverse set of tasks across multiple datasets, and show that ClinicalT5 dramatically outperforms T5 in the domain-specific tasks and compares favorably with its close baselines.
From a visual scene containing multiple people, human is able to distinguish each individual given the context descriptions about what happened before, their mental/physical states or intentions, etc. Above ability heavily relies on human-centric commonsense knowledge and reasoning. For example, if asked to identify the “person who needs healing” in an image, we need to first know that they usually have injuries or suffering expressions, then find the corresponding visual clues before finally grounding the person. We present a new commonsense task, Human-centric Commonsense Grounding, that tests the models’ ability to ground individuals given the context descriptions about what happened before, and their mental/physical states or intentions. We further create a benchmark, HumanCog, a dataset with 130k grounded commonsensical descriptions annotated on 67k images, covering diverse types of commonsense and visual scenes. We set up a context-object-aware method as a strong baseline that outperforms previous pre-trained and non-pretrained models. Further analysis demonstrates that rich visual commonsense and powerful integration of multi-modal commonsense are essential, which sheds light on future works. Data and code will be available at https://github.com/Hxyou/HumanCog.
Social media has increasingly played a key role in emergency response: first responders can use public posts to better react to ongoing crisis events and deploy the necessary resources where they are most needed. Timeline extraction and abstractive summarization are critical technical tasks to leverage large numbers of social media posts about events. Unfortunately, there are few datasets for benchmarking technical approaches for those tasks. This paper presents , the largest dataset of local crisis event timelines available to date. contains 1,000 crisis event timelines across four domains: wildfires, local fires, traffic, and storms. We built using a semi-automated cluster-then-refine approach to collect data from the public Twitter stream. Our initial experiments indicate a significant gap between the performance of strong baselines compared to the human performance on both tasks.Our dataset, code, and models are publicly available (https://github.com/CrisisLTLSum/CrisisTimelines).
Pre-trained multilingual language models show significant performance gains for zero-shot cross-lingual model transfer on a wide range of natural language understanding (NLU) tasks. Previously, for zero-shot cross-lingual evaluation, pre-trained models are only fine-tuned on English data and tested on a variety of target languages. In this paper, we do cross-lingualevaluation on various NLU tasks (sentence classification, sequence labeling, question answering) using prompt-tuning and compare it with fine-tuning. The results show that prompt tuning achieves much better cross-lingual transfer than fine-tuning across datasets, with only 0.1% to 0.3% tuned parameters. Additionally, we demonstrate through the analysis that prompt tuning can have better cross-lingual transfer-ability of representations on downstream tasks with better aligned decision boundaries.
This paper presents BERT-CTC, a novel formulation of end-to-end speech recognition that adapts BERT for connectionist temporal classification (CTC). Our formulation relaxes the conditional independence assumptions used in conventional CTC and incorporates linguistic knowledge through the explicit output dependency obtained by BERT contextual embedding. BERT-CTC attends to the full contexts of the input and hypothesized output sequences via the self-attention mechanism. This mechanism encourages a model to learn inner/inter-dependencies between the audio and token representations while maintaining CTC’s training efficiency. During inference, BERT-CTC combines a mask-predict algorithm with CTC decoding, which iteratively refines an output sequence. The experimental results reveal that BERT-CTC improves over conventional approaches across variations in speaking styles and languages. Finally, we show that the semantic representations in BERT-CTC are beneficial towards downstream spoken language understanding tasks.
One of the key challenges of automatic story generation is how to generate a long narrative that can maintain fluency, relevance, and coherence. Despite recent progress, current story generation systems still face the challenge of how to effectively capture contextual and event features, which has a profound impact on a model’s generation performance. To address these challenges, we present EtriCA, a novel neural generation model, which improves the relevance and coherence of the generated stories through residually mapping context features to event sequences with a cross-attention mechanism. Such a feature capturing mechanism allows our model to better exploit the logical relatedness between events when generating stories. Extensive experiments based on both automatic and human evaluations show that our model significantly outperforms state-of-the-art baselines, demonstrating the effectiveness of our model in leveraging context and event features.
Natural language interaction is a promising direction for democratizing 3D shape design. However, existing methods for text-driven 3D shape editing face challenges in producing decoupled, local edits to 3D shapes. We address this problem by learning disentangled latent representations that ground language in 3D geometry. To this end, we propose a complementary tool set including a novel network architecture, a disentanglement loss, and a new editing procedure. Additionally, to measure edit locality, we define a new metric that we call part-wise edit precision. We show that our method outperforms existing SOTA methods by 20% in terms of edit locality, and up to 6.6% in terms of language reference resolution accuracy. Human evaluations additionally show that compared to the existing SOTA, our method produces shape edits that are more local, more semantically accurate, and more visually obvious. Our work suggests that by solely disentangling language representations, downstream 3D shape editing can become more local to relevant parts, even if the model was never given explicit part-based supervision.
In this paper, we study trade-offs between efficiency, cost and accuracy when pre-training Transformer encoders with different pre-training objectives. For this purpose, we analyze features of common objectives and combine them to create new effective pre-training approaches. Specifically, we designed light token generators based on a straightforward statistical approach, which can replace ELECTRA computationally heavy generators, thus highly reducing cost. Our experiments also show that (i) there are more efficient alternatives to BERT’s MLM, and (ii) it is possible to efficiently pre-train Transformer-based models using lighter generators without a significant drop in performance.
To reduce the toxic degeneration in a pretrained Language Model (LM), previous work on Language Model detoxification has focused on reducing the toxicity of the generation itself (self-toxicity) without consideration of the context. As a result, a type of implicit offensive language where the generations support the offensive language in the context is ignored. Different from the LM controlling tasks in previous work, where the desired attributes are fixed for generation, the desired stance of the generation depends on the offensiveness of the context. Therefore, we propose a novel control method to do context-dependent detoxification with the stance taken into consideration. We introduce meta prefixes to learn the contextualized stance control strategy and to generate the stance control prefix according to the input context. The generated stance prefix is then combined with the toxicity control prefix to guide the response generation. Experimental results show that our proposed method can effectively learn the context-dependent stance control strategies while keeping a low self-toxicity of the underlying LM.
Subevent Relation Extraction (SRE) is a task in Information Extraction that aims to recognize spatial and temporal containment relations between event mentions in text. Recent methods have utilized pre-trained language models to represent input texts for SRE. However, a key issue in existing SRE methods is the employment of sequential order of words in texts to feed into representation learning methods, thus unable to explicitly focus on important context words and their interactions to enhance representations. In this work, we introduce a new method for SRE that learns to induce effective graph structures for input texts to boost representation learning. Our method features a word alignment framework with dependency paths and optimal transport to identify important context words to form effective graph structures for SRE. In addition, to enable SRE research on non-English languages, we present a new multilingual SRE dataset for five typologically different languages. Extensive experiments reveal the state-of-the-art performance for our method on different datasets and languages.
Most existing approaches for Knowledge Base Question Answering (KBQA) focus on a specific underlying knowledge base either because of inherent assumptions in the approach, or because evaluating it on a different knowledge base requires non-trivial changes. However, many popular knowledge bases share similarities in their underlying schemas that can be leveraged to facilitate generalization across knowledge bases. To achieve this generalization, we introduce a KBQA framework based on a 2-stage architecture that explicitly separates semantic parsing from the knowledge base interaction, facilitating transfer learning across datasets and knowledge graphs. We show that pretraining on datasets with a different underlying knowledge base can nevertheless provide significant performance gains and reduce sample complexity. Our approach achieves comparable or state-of-the-art performance for LC-QuAD (DBpedia), WebQSP (Freebase), SimpleQuestions (Wikidata) and MetaQA (Wikimovies-KG).
Online discussions are abundant with opinions towards a common topic, and identifying (dis)agreement between a pair of comments enables many opinion mining applications. Realizing the increasing needs to analyze opinions for emergent new topics that however tend to lack annotations, we present the first meta-learning approach for few-shot (dis)agreement identification that can be quickly applied to analyze opinions for new topics with few labeled instances. Furthermore, we enhance the meta-learner’s domain generalization ability from two perspectives. The first is domain-invariant regularization, where we design a lexicon-based regularization loss to enable the meta-learner to learn domain-invariant cues. The second is domain-aware augmentation, where we propose domain-aware task augmentation for meta-training to learn domain-specific expressions. In addition to using an existing dataset, we also evaluate our approach on two very recent new topics, mask mandate and COVID vaccine, using our newly annotated datasets containing 1.5k and 1.4k SubReddits comment pairs respectively. Extensive experiments on three domains/topics demonstrate the effectiveness of our meta-learning approach.
While collecting or generating more parallel data is necessary to improve machine translation (MT) in low-resource settings, we lack an understanding of how the limited amounts of existing data are actually used to help guide the collection of further resources. In this paper, we apply data cartography techniques (Swayamdipta et al., 2020) to characterize the contribution of training samples in two low-resource MT tasks (Swahili-English and Turkish-English) throughout the training of standard neural MT models. Our empirical study shows that, unlike in prior work for classification tasks, most samples contribute to model training in low-resource MT, albeit not uniformly throughout the training process. Furthermore, uni-dimensional characterizations of samples – e.g., based on dual cross-entropy or word frequency – do not suffice to characterize to what degree they are hard or easy to learn. Taken together, our results suggest that data augmentation strategies for low-resource MT would benefit from model-in-the-loop strategies to maximize improvements.
The task of context-dependent text-to-SQL aims to convert multi-turn user utterances to formal SQL queries. This is a challenging task due to both the scarcity of training data from which to learn complex contextual dependencies and to generalize to unseen databases. In this paper we explore augmenting the training datasets using self-play, which leverages contextual information to synthesize new interactions to adapt the model to new databases. We first design a SQL-to-text model conditioned on a sampled goal query, which represents a user’s intent, that then converses with a text-to-SQL semantic parser to generate new interactions. We then filter the synthesized interactions and retrain the models with the augmented data. We find that self-play improves the accuracy of a strong baseline on SParC and CoSQL, two widely used cross-domain text-to-SQL datasets. Our analysis shows that self-play simulates various conversational thematic relations, enhances cross-domain generalization and improves beam-search.
We explore the idea of compressing the prompts used to condition language models, and show that compressed prompts can retain a substantive amount of information about the original prompt. For severely compressed prompts, while fine-grained information is lost, abstract information and general sentiments can be retained with surprisingly few parameters, which can be useful in the context of decode-time algorithms for controllability and toxicity reduction. We find that some complex prompts can be effectively compressed into a single token to guide generation. We also show that compressed prompts are largely compositional, and can be constructed such that they can be used to control independent aspects of generated text.
While a substantial body of prior work has explored adversarial example generation for natural language understanding tasks, these examples are often unrealistic and diverge from the real-world data distributions. In this work, we introduce a two-stage adversarial example generation framework (NaturalAdversaries), for designing adversaries that are effective at fooling a given classifier and demonstrate natural-looking failure cases that could plausibly occur during in-the-wild deployment of the models. At the first stage a token attribution method is used to summarize a given classifier’s behavior as a function of the key tokens in the input. In the second stage a generative model is conditioned on the key tokens from the first stage. NaturalAdversaries is adaptable to both black-box and white-box adversarial attacks based on the level of access to the model parameters. Our results indicate these adversaries generalize across domains, and offer insights for future research on improving robustness of neural text classification models.
For years the model performance in machine learning obeyed a power-law relationship with the model size. For the consideration of parameter efficiency, recent studies focus on increasing model depth rather than width to achieve better performance. In this paper, we study how model width affects the Transformer model through a parameter-efficient multi-path structure. To better fuse features extracted from diff