We propose an autoregressive entity linking model, that is trained with two auxiliary tasks, and learns to re-rank generated samples at inference time. Our proposed novelties address two weaknesses in the literature. First, a recent method proposes to learn mention detection and then entity candidate selection, but relies on predefined sets of candidates. We use encoder-decoder autoregressive entity linking in order to bypass this need, and propose to train mention detection as an auxiliary task instead. Second, previous work suggests that re-ranking could help correct prediction errors. We add a new, auxiliary task, match prediction, to learn re-ranking. Without the use of a knowledge base or candidate sets, our model sets a new state of the art in two benchmark datasets of entity linking: COMETA in the biomedical domain, and AIDA-CoNLL in the news domain. We show through ablation studies that each of the two auxiliary tasks increases performance, and that re-ranking is an important factor to the increase. Finally, our low-resource experimental results suggest that performance on the main task benefits from the knowledge learned by the auxiliary tasks, and not just from the additional training data.
User language data can contain highly sensitive personal content. As such, it is imperative to offer users a strong and interpretable privacy guarantee when learning from their data. In this work we propose SentDP, pure local differential privacy at the sentence level for a single user document. We propose a novel technique, DeepCandidate, that combines concepts from robust statistics and language modeling to produce high (768) dimensional, general 𝜖-SentDP document embeddings. This guarantees that any single sentence in a document can be substituted with any other sentence while keeping the embedding 𝜖-indistinguishable. Our experiments indicate that these private document embeddings are useful for downstream tasks like sentiment analysis and topic classification and even outperform baseline methods with weaker guarantees like word-level Metric DP.
Current medical question answering systems have difficulty processing long, detailed and informally worded questions submitted by patients, called Consumer Health Questions (CHQs). To address this issue, we introduce a medical question understanding and answering system with knowledge grounding and semantic self-supervision. Our system is a pipeline that first summarizes a long, medical, user-written question, using a supervised summarization loss. Then, our system performs a two-step retrieval to return answers. The system first matches the summarized user question with an FAQ from a trusted medical knowledge base, and then retrieves a fixed number of relevant sentences from the corresponding answer document. In the absence of labels for question matching or answer relevance, we design 3 novel, self-supervised and semantically-guided losses. We evaluate our model against two strong retrieval-based question answering baselines. Evaluators ask their own questions and rate the answers retrieved by our baselines and own system according to their relevance. They find that our system retrieves more relevant answers, while achieving speeds 20 times faster. Our self-supervised losses also help the summarizer achieve higher scores in ROUGE, as well as in human evaluation metrics.
In this paper, we present our findings in the two subtasks of the 2022 NADI shared task. First, in the Arabic dialect identification subtask, we find that there is heavy class imbalance, and propose to address this issue using focal loss. Our experiments with the focusing hyperparameter confirm that focal loss improves performance. Second, in the Arabic tweet sentiment analysis subtask, we deal with a smaller dataset, where text includes both Arabic dialects and Modern Standard Arabic. We propose to use transfer learning from both pre-trained MSA language models and our own model from the first subtask. Our system ranks in the 5th and 7th best spots of the leaderboards of first and second subtasks respectively.
Understanding the intent of medical questions asked by patients, or Consumer Health Questions, is an essential skill for medical Conversational AI systems. We propose a novel data-augmented and simple joint learning approach combining question summarization and Recognizing Question Entailment (RQE) in the medical domain. Our data augmentation approach enables to use just one dataset for joint learning. We show improvements on both tasks across four biomedical datasets in accuracy (+8%), ROUGE-1 (+2.5%) and human evaluation scores. Human evaluation shows joint learning generates faithful and informative summaries. Finally, we release our code, the two question summarization datasets extracted from a large-scale medical dialogue dataset, as well as our augmented datasets.
Users of medical question answering systems often submit long and detailed questions, making it hard to achieve high recall in answer retrieval. To alleviate this problem, we propose a novel Multi-Task Learning (MTL) method with data augmentation for medical question understanding. We first establish an equivalence between the tasks of question summarization and Recognizing Question Entailment (RQE) using their definitions in the medical domain. Based on this equivalence, we propose a data augmentation algorithm to use just one dataset to optimize for both tasks, with a weighted MTL loss. We introduce gradually soft parameter-sharing: a constraint for decoder parameters to be close, that is gradually loosened as we move to the highest layer. We show through ablation studies that our proposed novelties improve performance. Our method outperforms existing MTL methods across 4 datasets of medical question pairs, in ROUGE scores, RQE accuracy and human evaluation. Finally, we show that our method fares better than single-task learning under 4 low-resource settings.
Syntactic structure is an important component of natural language text. Recent top-performing models in Answer Sentence Selection (AS2) use self-attention and transfer learning, but not syntactic structure. Tree structures have shown strong performance in tasks with sentence pair input like semantic relatedness. We investigate whether tree structures can boost performance in AS2. We introduce the Tree Aggregation Transformer: a novel recursive, tree-structured self-attention model for AS2. The recursive nature of our model is able to represent all levels of syntactic parse trees with only one additional self-attention layer. Without transfer learning, we establish a new state of the art on the popular TrecQA and WikiQA benchmark datasets. Additionally, we evaluate our method on four Community Question Answering datasets, and find that tree-structured representations have limitations with noisy user-generated text. We conduct probing experiments to evaluate how our models leverage tree structures across datasets. Our findings show that the ability of tree-structured models to successfully absorb syntactic information is strongly correlated with a higher performance in AS2.
We consider the problem of topic-focused abstractive summarization, where the goal is to generate an abstractive summary focused on a particular topic, a phrase of one or multiple words. We hypothesize that the task of generating topic-focused summaries can be improved by showing the model what it must not focus on. We introduce a deep reinforcement learning approach to topic-focused abstractive summarization, trained on rewards with a novel negative example baseline. We define the input in this problem as the source text preceded by the topic. We adapt the CNN-Daily Mail and New York Times summarization datasets for this task. We then show through experiments on existing rewards that the use of a negative example baseline can outperform the use of a self-critical baseline, in Rouge, BERTScore, and human evaluation metrics.
In this paper, we describe our approach to question summarization and multi-answer summarization in the context of the 2021 MEDIQA shared task (Ben Abacha et al., 2021). We propose two kinds of transfer learning for the abstractive summarization of medical questions. First, we train on HealthCareMagic, a large question summarization dataset collected from an online healthcare service platform. Second, we leverage the ability of the BART encoder-decoder architecture to model both generation and classification tasks to train on the task of Recognizing Question Entailment (RQE) in the medical domain. We show that both transfer learning methods combined achieve the highest ROUGE scores. Finally, we cast the question-driven extractive summarization of multiple relevant answer documents as an Answer Sentence Selection (AS2) problem. We show how we can preprocess the MEDIQA-AnS dataset such that it can be trained in an AS2 setting. Our AS2 model is able to generate extractive summaries achieving high ROUGE scores.
Attention mechanisms have improved the performance of NLP tasks while allowing models to remain explainable. Self-attention is currently widely used, however interpretability is difficult due to the numerous attention distributions. Recent work has shown that model representations can benefit from label-specific information, while facilitating interpretation of predictions. We introduce the Label Attention Layer: a new form of self-attention where attention heads represent labels. We test our novel layer by running constituency and dependency parsing experiments and show our new model obtains new state-of-the-art results for both tasks on both the Penn Treebank (PTB) and Chinese Treebank. Additionally, our model requires fewer self-attention layers compared to existing work. Finally, we find that the Label Attention heads learn relations between syntactic categories and show pathways to analyze errors.
Moroccan Darija is a variant of Arabic with many influences. Using the Open Multilingual WordNet (OMW), we compare the lemmas in the Moroccan Darija Wordnet (MDW) with the standard Arabic, French and Spanish ones. We then compared the lemmas in each synset with their translation equivalents. Transliteration is used to bridge alphabet differences and match lemmas in the closest phonological way. The results put figures on the similarity Moroccan Darija has with Arabic, French and Spanish: respectively 42.0%, 2.8% and 2.2%.
We propose methods to link automatically parsed linguistic data to the WordNet. We apply these methods on a trilingual dictionary in Fula, English and French. Dictionary entry parsing is used to collect the linguistic data. Then we connect it to the Open Multilingual WordNet (OMW) through two attempts, and use confidence scores to quantify accuracy. We obtained 11,000 entries in parsing and linked about 58% to the OMW on the first attempt, and an additional 14% in the second one. These links are due to be validated by Fula speakers before being added to the Kamusi Project’s database.