Bilateral Multi-Perspective Matching for Natural Language Sentences

Abstract

Natural language sentence matching is a fundamental technology for a variety of tasks. Previous approaches either match sentences from a single direction or only apply single granular (word-by-word or sentence-by-sentence) matching. In this work, we propose a bilateral multi-perspective matching (BiMPM) model under the "matching-aggregation" framework. Given two sentences $P$ and $Q$, our model first encodes them with a BiLSTM encoder. Next, we match the two encoded sentences in two directions $P\to Q$ and $P\leftarrow Q$. In each matching direction, each time step of one sentence is matched against all time-steps of the other sentence from multiple perspectives. Then, another BiLSTM layer is utilized to aggregate the matching results into a fix-length matching vector. Finally, based on the matching vector, the decision is made through a fully connected layer. We evaluate our model on three tasks: paraphrase identification, natural language inference and answer sentence selection. Experimental results on standard benchmark datasets show that our model achieves the state-of-the-art performance on all tasks.

arXiv commentTo appear in Proceedings of IJCAI 2017

References in corpus (17)

• Adam: A Method for Stochastic Optimization
• Enhanced LSTM for Natural Language Inference
• Reasoning about Entailment with Neural Attention
• A large annotated corpus for learning natural language inference
• LSTM-based Deep Learning Models for Non-factoid Answer Selection
• Attentive Pooling Networks
• Learning Natural Language Inference using Bidirectional LSTM model and Inner-Attention
• Long Short-Term Memory-Networks for Machine Reading
• Sentence Similarity Learning by Lexical Decomposition and Composition
• Multi-Perspective Context Matching for Machine Comprehension
• Order-Embeddings of Images and Language
• Learning Natural Language Inference with LSTM
• A Decomposable Attention Model for Natural Language Inference
• ABCNN: Attention-Based Convolutional Neural Network for Modeling Sentence Pairs
• A Recurrent Encoder-Decoder Network for Sequential Face Alignment
• Modelling Interaction of Sentence Pair with coupled-LSTMs
• Semi-supervised Clustering for Short Text via Deep Representation Learning

Cited by in corpus (63)

• The Natural Language Decathlon: Multitask Learning as Question Answering
• Neural Natural Language Inference Models Enhanced with External Knowledge
• Pathologies of Neural Models Make Interpretations Difficult
• Natural Language Inference over Interaction Space
• A Deep Network Model for Paraphrase Detection in Short Text Messages
• Neural Network Models for Paraphrase Identification, Semantic Textual Similarity, Natural Language Inference, and Question Answering
• A Survey on Text Classification: From Shallow to Deep Learning
• Learning General Purpose Distributed Sentence Representations via Large Scale Multi-task Learning
• Improved Neural Relation Detection for Knowledge Base Question Answering
• Stochastic Answer Networks for Natural Language Inference
• Semantic Sentence Matching with Densely-connected Recurrent and Co-attentive Information
• Multiresolution Graph Attention Networks for Relevance Matching
• Shortcut-Stacked Sentence Encoders for Multi-Domain Inference
• Enhance word representation for out-of-vocabulary on Ubuntu dialogue corpus
• Neural Matching Models for Question Retrieval and Next Question Prediction in Conversation
• Learning Loss Functions for Semi-supervised Learning via Discriminative Adversarial Networks
• Multi-Cast Attention Networks for Retrieval-based Question Answering and Response Prediction
• Dynamic Fusion Networks for Machine Reading Comprehension
• Match-Ignition: Plugging PageRank into Transformer for Long-form Text Matching
• Question Answering through Transfer Learning from Large Fine-grained Supervision Data
• Lightweight and Efficient Neural Natural Language Processing with Quaternion Networks
• Speaker-Aware BERT for Multi-Turn Response Selection in Retrieval-Based Chatbots
• Improving Natural Language Inference Using External Knowledge in the Science Questions Domain
• To Tune or Not To Tune? How About the Best of Both Worlds?
• Explicit Interaction Model towards Text Classification
• Improving Neural Question Generation using Answer Separation
• Neural Paraphrase Identification of Questions with Noisy Pretraining
• The Inception Team at NSURL-2019 Task 8: Semantic Question Similarity in Arabic
• Multi-task Learning for Universal Sentence Embeddings: A Thorough Evaluation using Transfer and Auxiliary Tasks
• LadRa-Net: Locally-Aware Dynamic Re-read Attention Net for Sentence Semantic Matching
• Building Sequential Inference Models for End-to-End Response Selection
• Identifying Well-formed Natural Language Questions
• Lattice CNNs for Matching Based Chinese Question Answering
• Adversarial Training for Community Question Answer Selection Based on Multi-scale Matching
• Stance Prediction for Russian: Data and Analysis
• Attention Boosted Sequential Inference Model
• BanditRank: Learning to Rank Using Contextual Bandits
• Sentence Encoding with Tree-constrained Relation Networks
• Learning Robust, Transferable Sentence Representations for Text Classification
• Improved Semantic-Aware Network Embedding with Fine-Grained Word Alignment
• Attentive Convolution: Equipping CNNs with RNN-style Attention Mechanisms
• Explaining Neural Network Predictions on Sentence Pairs via Learning Word-Group Masks
• Retrieve Synonymous keywords for Frequent Queries in Sponsored Search in a Data Augmentation Way
• Match${}^2$: A Matching over Matching Model for Similar Question Identification
• Toward the Understanding of Deep Text Matching Models for Information Retrieval
• Wasserstein Distance Regularized Sequence Representation for Text Matching in Asymmetrical Domains
• AWE: Asymmetric Word Embedding for Textual Entailment
• On Tree-Based Neural Sentence Modeling
• On Generality and Knowledge Transferability in Cross-Domain Duplicate Question Detection for Heterogeneous Community Question Answering
• End-Task Oriented Textual Entailment via Deep Explorations of Inter-Sentence Interactions
• Combining word embeddings and convolutional neural networks to detect duplicated questions
• Hierarchical Gated Recurrent Neural Tensor Network for Answer Triggering
• Unsupervised Question Answering for Fact-Checking
• REGMAPR - Text Matching Made Easy
• Utterance-to-Utterance Interactive Matching Network for Multi-Turn Response Selection in Retrieval-Based Chatbots
• A multi-perspective combined recall and rank framework for Chinese procedure terminology normalization
• Dropping Networks for Transfer Learning
• EPIHC: Improving Enhancer-Promoter Interaction Prediction by using Hybrid features and Communicative learning
• DGA-Net Dynamic Gaussian Attention Network for Sentence Semantic Matching
• Contextualized Embeddings based Convolutional Neural Networks for Duplicate Question Identification
• Pre-trained Language Model Based Active Learning for Sentence Matching
• Convexification of Neural Graph
• FILM: A Fast, Interpretable, and Low-rank Metric Learning Approach for Sentence Matching