Natural Language Generation (NLG) typically involves evaluating the generated text in various aspects (e.g., consistency and naturalness) to obtain a comprehensive assessment. However, multi-aspect evaluation remains challenging as it may require the evaluator to generalize to any given evaluation aspect even if it’s absent during training. In this paper, we introduce X-Eval, a two-stage instruction tuning framework to evaluate text in both seen and unseen aspects customized by end users. X-Eval consists of two learning stages: the vanilla instruction tuning stage that improves the model’s ability to follow evaluation instructions, and an enhanced instruction tuning stage that exploits the connections between fine-grained evaluation aspects to better assess text quality. To support the training of X-Eval, we collect AspectInstruct, the first instruction tuning dataset tailored for multi-aspect NLG evaluation spanning 27 diverse evaluation aspects with 65 tasks. To enhance task diversity, we devise an augmentation strategy that converts human rating annotations into diverse forms of NLG evaluation tasks, including scoring, comparison, ranking, and Boolean question answering. Extensive experiments across three essential categories of NLG tasks: dialogue generation, summarization, and data-to-text coupled with 21 aspects in meta-evaluation, demonstrate that X-Eval enables even a lightweight language model to achieve a comparable if not higher correlation with human judgments compared to the state-of-the-art NLG evaluators like GPT-4.
Despite vision-language models’ (VLMs) remarkable capabilities as versatile visual assistants, two substantial challenges persist within the existing VLM frameworks: (1) lacking task diversity in pretraining and visual instruction tuning, and (2) annotation error and bias in GPT-4 synthesized instruction tuning data. Both challenges lead to issues such as poor generalizability, hallucination, and catastrophic forgetting. To address these challenges, we construct Vision-Flan, the most diverse publicly available visual instruction tuning dataset to date, comprising 187 diverse tasks and 1,664,261 instances sourced from academic datasets, and each task is accompanied by an expert-written instruction. In addition, we propose a two-stage instruction tuning framework, in which VLMs are firstly finetuned on Vision-Flan and further tuned on GPT-4 synthesized data. We find this two-stage tuning framework significantly outperforms the traditional single-stage visual instruction tuning framework and achieves the state-of-the-art performance across a wide range of multi-modal evaluation benchmarks. Finally, we conduct in-depth analyses to understand visual instruction tuning and our findings reveal that: (1) GPT-4 synthesized data does not substantially enhance VLMs’ capabilities but rather modulates the model’s responses to human-preferred formats; (2) A minimal quantity (e.g., 1,000) of GPT-4 synthesized data can effectively align VLM responses with human-preference; (3) Visual instruction tuning mainly helps large-language models (LLMs) to understand visual features.
Data visualization serves as a critical means for presenting data and mining its valuable insights. The task of chart summarization, through natural language processing techniques, facilitates in-depth data analysis of charts. However, there still are notable deficiencies in terms of visual-language matching and reasoning ability for existing approaches. To address these limitations, this study constructs a large-scale dataset of comprehensive chart-caption pairs and fine-tuning instructions on each chart. Thanks to the broad coverage of various topics and visual styles within this dataset, better matching degree can be achieved from the view of training data. Moreover, we propose an innovative chart summarization method, ChartThinker, which synthesizes deep analysis based on chains of thought and strategies of context retrieval, aiming to improve the logical coherence and accuracy of the generated summaries. Built upon the curated datasets, our trained model consistently exhibits superior performance in chart summarization tasks, surpassing 8 state-of-the-art models over 7 evaluation metrics. Our dataset and codes are publicly accessible.
Aspect-Based Sentiment Analysis (ABSA) aims to determine the sentiment polarities of specified aspect terms in a sentence. Most previous approaches mainly use an attention mechanism or graph neural networks based on dependency trees to explicitly model the connections between aspect terms and opinion words. However, these methods may not effectively address cases where the sentiment of an aspect term is implicitly described, as the corresponding opinion words may not directly appear in the sentence. To alleviate this issue, in this paper, we propose a GCNet that explicitly leverages global semantic information to guide context encoding. Particularly, we design a semantics encoding module that incorporates global semantic features into sequential modeling process to enable the consideration of the overall sentiment tendency of a sentence, while the global semantic features are also refined by adaptively focusing on different parts of the sentence. Moreover, for a comprehensive sentence analysis, we also include a syntactic feature encoding module along with a pre-fusion module to integrate the refined global features with the syntactic representations. Extensive experiments on three public datasets demonstrate that our model outperforms state-of-the-art methods, indicating the robustness and effectiveness of our approach.
Unsupervised Domain Adaptation (UDA) of the Aspect-based Sentiment Analysis (ABSA) task aims to transfer knowledge learned from labeled source domain datasets to unlabeled target domains on the assumption that samples from the source domain are freely accessible during the training period. However, this assumption can easily lead to privacy invasion issues in real-world applications, especially when the source data involves privacy-preserving domains such as healthcare and finance. In this paper, we introduce the Source-Free Domain Adaptation Framework for ABSA (SF-ABSA), which only allows model parameter transfer, not data transfer, between different domains. Specifically, the proposed SF-ABSA framework consists of two parts, i.e., feature-based adaptation and pseudo-label-based adaptation. Experiment results on four benchmarks show that the proposed framework performs competitively with traditional unsupervised domain adaptation methods under the premise of insufficient information, which demonstrates the superiority of our method under privacy conditions.
Multimodal Large Language Models (MLLMs) have demonstrated remarkable proficiency in diverse tasks across different domains, with an increasing focus on improving their zero-shot generalization capabilities for unseen multimodal tasks. Multimodal instruction tuning has emerged as a successful strategy for achieving zero-shot generalization by fine-tuning pre-trained models on diverse multimodal tasks through instructions. As MLLMs grow in complexity and size, the need for parameter-efficient fine-tuning methods like Low-Rank Adaption (LoRA), which fine-tunes with a minimal set of parameters, becomes essential. However, applying LoRA in multimodal instruction tuning presents the challenge of task interference, which leads to performance degradation, especially when dealing with a broad array of multimodal tasks. To address this, this paper introduces a novel approach that integrates multimodal instruction tuning with Conditional Mixture-of-LoRA (MixLoRA). It innovates upon LoRA by dynamically constructing low-rank adaptation matrices tailored to the unique demands of each input instance, aiming to mitigate task interference. Experimental results on various multimodal evaluation datasets indicate that MixLoRA not only outperforms the conventional LoRA with the same or even higher ranks, demonstrating its efficacy and adaptability in diverse multimodal tasks.
Multiple-choice questions (MCQs) are important in enhancing concept learning and student engagement for educational purposes. Despite the multimodal nature of educational content, current methods focus mainly on text-based inputs and often neglect the integration of visual information. In this work, we study the problem of multimodal educational question generation, which aims at generating subject-specific educational questions with plausible yet incorrect distractors based on multimodal educational content. To tackle this problem, we introduce a novel framework, named Chain-of-Exemplar (CoE), which utilizes multimodal large language models (MLLMs) with Chain-of-Thought reasoning to improve the generation of challenging distractors. Furthermore, CoE leverages three-stage contextualized exemplar retrieval to retrieve exemplary questions as guides for generating more subject-specific educational questions. Experimental results on the ScienceQA benchmark demonstrate the superiority of CoE in both question generation and distractor generation over existing methods across various subjects and educational levels.
Writing assistance aims to improve the correctness and quality of input texts, with character checking being crucial in detecting and correcting wrong characters. In the real world where handwriting occupies the vast majority, characters that humans get wrong include faked characters (i.e., untrue characters created due to writing errors) and misspelled characters (i.e., true characters used incorrectly due to spelling errors). However, existing datasets and related studies only focus on misspelled characters that can be represented by computer text encoding systems, thereby ignoring faked characters which are more common and difficult. To break through this dilemma, we present Visual-C3, a human-annotated VisualChinese Character Checking dataset with faked and misspelled Chinese characters. To the best of our knowledge, Visual-C3 is the first real-world visual and the largest human-crafted dataset for the Chinese character checking scenario. Additionally, we also propose and evaluate novel baseline methods on Visual-C3. Extensive empirical results and analyses show that Visual-C3 is high-quality yet challenging. As the first study focusing on Chinese faked characters, the dataset and the baseline methods are publicly available at https://github.com/THUKElab/Visual-C3.
Multimodal question answering (MMQA), which aims to derive the answer from multiple knowledge modalities (e.g., text, tables, and images), has received increasing attention due to its board applications. Current approaches to MMQA often rely on single-modal or bi-modal QA models, which limits their ability to effectively integrate information across all modalities and leverage the power of pre-trained language models. To address these limitations, we propose a novel framework called UniMMQA, which unifies three different input modalities into a text-to-text format by employing position-enhanced table linearization and diversified image captioning techniques. Additionally, we enhance cross-modal reasoning by incorporating a multimodal rationale generator, which produces textual descriptions of cross-modal relations for adaptation into the text-to-text generation process. Experimental results on three MMQA benchmark datasets show the superiority of UniMMQA in both supervised and unsupervised settings.
Federated learning (FL) enables multiple participants to collaboratively train machine learning models using decentralized data sources, alleviating privacy concerns that arise from directly sharing local data. However, the lack of model privacy protection in FL becomes an unneglectable challenge, especially when people want to federally finetune models based on a proprietary large language model. In this study, we propose a novel FL training approach that accomplishes information exchange among participants via tunable soft prompts. These soft prompts, updated and transmitted between the server and clients, assume the role of the global model parameters and serve as messengers to deliver useful knowledge from the local data and global model. As the global model itself is not required to be shared and the local training is conducted based on an auxiliary model with fewer parameters than the global model, the proposed approach provides protection for the global model while reducing communication and computation costs in FL. Extensive experiments show the effectiveness of the proposed approach compared to several baselines. We have released the source code at https://github.com/alibaba/FederatedScope/tree/fedsp/federatedscope/nlp/fedsp.
Chain-of-Thought (CoT) prompting enables large language models to solve complex reasoning problems by generating intermediate steps. However, confined by its inherent single-pass and sequential generation process, CoT heavily relies on the initial decisions, causing errors in early steps to accumulate and impact the final answers. In contrast, humans adopt recursive thinking when tackling complex reasoning problems, i.e. iteratively breaking the original problem into approachable sub-problems and aggregating their answers to resolve the original one. Inspired by the human cognitive process, we propose SOCRATIC QUESTIONING, a divide-and-conquer style algorithm that mimics the recursive thinking process. Specifically, SOCRATIC QUESTIONING leverages large language models to raise and answer sub-questions until collecting enough information to tackle the original question. Unlike CoT, SOCRATIC QUESTIONING explicitly navigates the thinking space, stimulates effective recursive thinking, and is more robust towards errors in the thinking process. Extensive experiments on several complex reasoning tasks, including MMLU, MATH, LogiQA, and visual question-answering demonstrate significant performance improvements over the state-of-the-art prompting methods, such as CoT, and Tree-of-Thought. The qualitative analysis clearly shows that the intermediate reasoning steps elicited by SOCRATIC QUESTIONING are similar to humans’ recursively thinking process of complex reasoning problems.
Evaluating the performance of Grammatical Error Correction (GEC) systems is a challenging task due to its subjectivity. Designing an evaluation metric that is as objective as possible is crucial to the development of GEC task. However, mainstream evaluation metrics, i.e., reference-based metrics, introduce bias into the multi-reference evaluation by extracting edits without considering the presence of multiple references. To overcome this issue, we propose Chunk-LE Multi-reference Evaluation (CLEME), designed to evaluate GEC systems in the multi-reference evaluation setting. CLEME builds chunk sequences with consistent boundaries for the source, the hypothesis and references, thus eliminating the bias caused by inconsistent edit boundaries. Furthermore, we observe the consistent boundary could also act as the boundary of grammatical errors, based on which the F0.5 score is then computed following the correction independence assumption. We conduct experiments on six English reference sets based on the CoNLL-2014 shared task. Extensive experiments and detailed analyses demonstrate the correctness of our discovery and the effectiveness of CLEME. Further analysis reveals that CLEME is robust to evaluate GEC systems across reference sets with varying numbers of references and annotation styles. All the source codes of CLEME are released at https://github.com/THUKElab/CLEME.
Instruction tuning, a new learning paradigm that fine-tunes pre-trained language models on tasks specified through instructions, has shown promising zero-shot performance on various natural language processing tasks. However, it has yet to be explored for vision and multimodal tasks. In this work, we introduce MultiInstruct, the first multimodal instruction tuning benchmark dataset that consists of 62 diverse multimodal tasks in a unified seq-to-seq format covering 10 broad categories. The tasks are derived from 21 existing open-source datasets and each task is equipped with 5 expert-written instructions. We take OFA as the base pre-trained model for multimodal instruction tuning, and to further improve its zero-shot performance, we explore multiple transfer learning strategies to leverage the large-scale Natural Instructions dataset. Experimental results demonstrate strong zero-shot performance on various unseen multimodal tasks and the benefit of transfer learning from a text-only instruction dataset. We also design a new evaluation metric – Sensitivity, to evaluate how sensitive the model is to the variety of instructions. Our results indicate that fine-tuning the model on a diverse set of tasks and instructions leads to a reduced sensitivity to variations in instructions for each task.
Controllable Text Generation (CTG) has obtained great success due to its fine-grained generation ability obtained by focusing on multiple attributes. However, most existing CTG researches overlook how to utilize the attribute entanglement to enhance the diversity of the controlled generated texts. Facing this dilemma, we focus on a novel CTG scenario, i.e., blessing generation which is challenging because high-quality blessing texts require CTG models to comprehensively consider the entanglement between multiple attributes (e.g., objects and occasions). To promote the research on blessing generation, we present EBleT, a large-scale Entangled Blessing Text dataset containing 293K English sentences annotated with multiple attributes. Furthermore, we propose novel evaluation metrics to measure the quality of the blessing texts generated by the baseline models we designed. Our study opens a new research direction for controllable text generation and enables the development of attribute-entangled CTG models.
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.
On many natural language processing tasks, large pre-trained language models (PLMs) have shown overwhelming performances compared with traditional neural network methods. Nevertheless, their huge model size and low inference speed have hindered the deployment on resource-limited devices in practice. In this paper, we target to compress PLMs with knowledge distillation, and propose a hierarchical relational knowledge distillation (HRKD) method to capture both hierarchical and domain relational information. Specifically, to enhance the model capability and transferability, we leverage the idea of meta-learning and set up domain-relational graphs to capture the relational information across different domains. And to dynamically select the most representative prototypes for each domain, we propose a hierarchical compare-aggregate mechanism to capture hierarchical relationships. Extensive experiments on public multi-domain datasets demonstrate the superior performance of our HRKD method as well as its strong few-shot learning ability. For reproducibility, we release the code at https://github.com/cheneydon/hrkd.
Transfer learning (TL) seeks to improve the learning of a data-scarce target domain by using information from source domains. However, the source and target domains usually have different data distributions, which may lead to negative transfer. To alleviate this issue, we propose a Wasserstein Selective Transfer Learning (WSTL) method. Specifically, the proposed method considers a reinforced selector to select helpful data for transfer learning. We further use a Wasserstein-based discriminator to maximize the empirical distance between the selected source data and target data. The TL module is then trained to minimize the estimated Wasserstein distance in an adversarial manner and provides domain invariant features for the reinforced selector. We adopt an evaluation metric based on the performance of the TL module as delayed reward and a Wasserstein-based metric as immediate rewards to guide the reinforced selector learning. Compared with the competing TL approaches, the proposed method selects data samples that are closer to the target domain. It also provides better state features and reward signals that lead to better performance with faster convergence. Extensive experiments on three real-world text mining tasks demonstrate the effectiveness of the proposed method.
This ability to learn consecutive tasks without forgetting how to perform previously trained problems is essential for developing an online dialogue system. This paper proposes an effective continual learning method for the task-oriented dialogue system with iterative network pruning, expanding, and masking (TPEM), which preserves performance on previously encountered tasks while accelerating learning progress on subsequent tasks. Specifically, TPEM (i) leverages network pruning to keep the knowledge for old tasks, (ii) adopts network expanding to create free weights for new tasks, and (iii) introduces task-specific network masking to alleviate the negative impact of fixed weights of old tasks on new tasks. We conduct extensive experiments on seven different tasks from three benchmark datasets and show empirically that TPEM leads to significantly improved results over the strong competitors.
Conversational Emotion Recognition (CER) is a crucial task in Natural Language Processing (NLP) with wide applications. Prior works in CER generally focus on modeling emotion influences solely with utterance-level features, with little attention paid on phrase-level semantic connection between utterances. Phrases carry sentiments when they are referred to emotional events under certain topics, providing a global semantic connection between utterances throughout the entire conversation. In this work, we propose a two-stage Summarization and Aggregation Graph Inference Network (SumAggGIN), which seamlessly integrates inference for topic-related emotional phrases and local dependency reasoning over neighbouring utterances in a global-to-local fashion. Topic-related emotional phrases, which constitutes the global topic-related emotional connections, are recognized by our proposed heterogeneous Summarization Graph. Local dependencies, which captures short-term emotional effects between neighbouring utterances, are further injected via an Aggregation Graph to distinguish the subtle differences between utterances containing emotional phrases. The two steps of graph inference are tightly-coupled for a comprehensively understanding of emotional fluctuation. Experimental results on three CER benchmark datasets verify the effectiveness of our proposed model, which outperforms the state-of-the-art approaches.
Dialogue Act Recognition (DAR) is a challenging problem in Natural Language Understanding, which aims to attach Dialogue Act (DA) labels to each utterance in a conversation. However, previous studies cannot fully recognize the specific expressions given by users due to the informality and diversity of natural language expressions. To solve this problem, we propose a Heterogeneous User History (HUH) graph convolution network, which utilizes the user’s historical answers grouped by DA labels as additional clues to recognize the DA label of utterances. To handle the noise caused by introducing the user’s historical answers, we design sets of denoising mechanisms, including a History Selection process, a Similarity Re-weighting process, and an Edge Re-weighting process. We evaluate the proposed method on two benchmark datasets MSDialog and MRDA. The experimental results verify the effectiveness of integrating user’s historical answers, and show that our proposed model outperforms the state-of-the-art methods.
Deep question generation (DQG) aims to generate complex questions through reasoning over multiple documents. The task is challenging and underexplored. Existing methods mainly focus on enhancing document representations, with little attention paid to the answer information, which may result in the generated question not matching the answer type and being answerirrelevant. In this paper, we propose an Answer-driven Deep Question Generation (ADDQG) model based on the encoder-decoder framework. The model makes better use of the target answer as a guidance to facilitate question generation. First, we propose an answer-aware initialization module with a gated connection layer which introduces both document and answer information to the decoder, thus helping to guide the choice of answer-focused question words. Then a semantic-rich fusion attention mechanism is designed to support the decoding process, which integrates the answer with the document representations to promote the proper handling of answer information during generation. Moreover, reinforcement learning is applied to integrate both syntactic and semantic metrics as the reward to enhance the training of the ADDQG. Extensive experiments on the HotpotQA dataset show that ADDQG outperforms state-of-the-art models in both automatic and human evaluations.
Distant supervision based methods for entity and relation extraction have received increasing popularity due to the fact that these methods require light human annotation efforts. In this paper, we consider the problem of shifted label distribution, which is caused by the inconsistency between the noisy-labeled training set subject to external knowledge graph and the human-annotated test set, and exacerbated by the pipelined entity-then-relation extraction manner with noise propagation. We propose a joint extraction approach to address this problem by re-labeling noisy instances with a group of cooperative multiagents. To handle noisy instances in a fine-grained manner, each agent in the cooperative group evaluates the instance by calculating a continuous confidence score from its own perspective; To leverage the correlations between these two extraction tasks, a confidence consensus module is designed to gather the wisdom of all agents and re-distribute the noisy training set with confidence-scored labels. Further, the confidences are used to adjust the training losses of extractors. Experimental results on two real-world datasets verify the benefits of re-labeling noisy instance, and show that the proposed model significantly outperforms the state-of-the-art entity and relation extraction methods.
The challenge of both achieving task completion by querying the knowledge base and generating human-like responses for task-oriented dialogue systems is attracting increasing research attention. In this paper, we propose a “Two-Teacher One-Student” learning framework (TTOS) for task-oriented dialogue, with the goal of retrieving accurate KB entities and generating human-like responses simultaneously. TTOS amalgamates knowledge from two teacher networks that together provide comprehensive guidance to build a high-quality task-oriented dialogue system (student network). Each teacher network is trained via reinforcement learning with a goal-specific reward, which can be viewed as an expert towards the goal and transfers the professional characteristic to the student network. Instead of adopting the classic student-teacher learning of forcing the output of a student network to exactly mimic the soft targets produced by the teacher networks, we introduce two discriminators as in generative adversarial network (GAN) to transfer knowledge from two teachers to the student. The usage of discriminators relaxes the rigid coupling between the student and teachers. Extensive experiments on two benchmark datasets (i.e., CamRest and In-Car Assistant) demonstrate that TTOS significantly outperforms baseline methods.
Chinese relation extraction is conducted using neural networks with either character-based or word-based inputs, and most existing methods typically suffer from segmentation errors and ambiguity of polysemy. To address the issues, we propose a multi-grained lattice framework (MG lattice) for Chinese relation extraction to take advantage of multi-grained language information and external linguistic knowledge. In this framework, (1) we incorporate word-level information into character sequence inputs so that segmentation errors can be avoided. (2) We also model multiple senses of polysemous words with the help of external linguistic knowledge, so as to alleviate polysemy ambiguity. Experiments on three real-world datasets in distinct domains show consistent and significant superiority and robustness of our model, as compared with other baselines. We will release the source code of this paper in the future.
Distantly supervised relation extraction greatly reduces human efforts in extracting relational facts from unstructured texts. However, it suffers from noisy labeling problem, which can degrade its performance. Meanwhile, the useful information expressed in knowledge graph is still underutilized in the state-of-the-art methods for distantly supervised relation extraction. In the light of these challenges, we propose CORD, a novelCOopeRativeDenoising framework, which consists two base networks leveraging text corpus and knowledge graph respectively, and a cooperative module involving their mutual learning by the adaptive bi-directional knowledge distillation and dynamic ensemble with noisy-varying instances. Experimental results on a real-world dataset demonstrate that the proposed method reduces the noisy labels and achieves substantial improvement over the state-of-the-art methods.
Review text has been widely studied in traditional tasks such as sentiment analysis and aspect extraction. However, to date, no work is towards the abstractive review summarization that is essential for business organizations and individual consumers to make informed decisions. This work takes the lead to study the aspect/sentiment-aware abstractive review summarization by exploring multi-factor attentions. Specifically, we propose an interactive attention mechanism to interactively learns the representations of context words, sentiment words and aspect words within the reviews, acted as an encoder. The learned sentiment and aspect representations are incorporated into the decoder to generate aspect/sentiment-aware review summaries via an attention fusion network. In addition, the abstractive summarizer is jointly trained with the text categorization task, which helps learn a category-specific text encoder, locating salient aspect information and exploring the variations of style and wording of content with respect to different text categories. The experimental results on a real-life dataset demonstrate that our model achieves impressive results compared to other strong competitors.
Answer selection is an important but challenging task. Significant progresses have been made in domains where a large amount of labeled training data is available. However, obtaining rich annotated data is a time-consuming and expensive process, creating a substantial barrier for applying answer selection models to a new domain which has limited labeled data. In this paper, we propose Knowledge-aware Attentive Network (KAN), a transfer learning framework for cross-domain answer selection, which uses the knowledge base as a bridge to enable knowledge transfer from the source domain to the target domains. Specifically, we design a knowledge module to integrate the knowledge-based representational learning into answer selection models. The learned knowledge-based representations are shared by source and target domains, which not only leverages large amounts of cross-domain data, but also benefits from a regularization effect that leads to more general representations to help tasks in new domains. To verify the effectiveness of our model, we use SQuAD-T dataset as the source domain and three other datasets (i.e., Yahoo QA, TREC QA and InsuranceQA) as the target domains. The experimental results demonstrate that KAN has remarkable applicability and generality, and consistently outperforms the strong competitors by a noticeable margin for cross-domain answer selection.
Multimodal research is an emerging field of artificial intelligence, and one of the main research problems in this field is multimodal fusion. The fusion of multimodal data is the process of integrating multiple unimodal representations into one compact multimodal representation. Previous research in this field has exploited the expressiveness of tensors for multimodal representation. However, these methods often suffer from exponential increase in dimensions and in computational complexity introduced by transformation of input into tensor. In this paper, we propose the Low-rank Multimodal Fusion method, which performs multimodal fusion using low-rank tensors to improve efficiency. We evaluate our model on three different tasks: multimodal sentiment analysis, speaker trait analysis, and emotion recognition. Our model achieves competitive results on all these tasks while drastically reducing computational complexity. Additional experiments also show that our model can perform robustly for a wide range of low-rank settings, and is indeed much more efficient in both training and inference compared to other methods that utilize tensor representations.