Sentence Similarity Learning by Lexical Decomposition and Composition

Abstract

Most conventional sentence similarity methods only focus on similar parts of two input sentences, and simply ignore the dissimilar parts, which usually give us some clues and semantic meanings about the sentences. In this work, we propose a model to take into account both the similarities and dissimilarities by decomposing and composing lexical semantics over sentences. The model represents each word as a vector, and calculates a semantic matching vector for each word based on all words in the other sentence. Then, each word vector is decomposed into a similar component and a dissimilar component based on the semantic matching vector. After this, a two-channel CNN model is employed to capture features by composing the similar and dissimilar components. Finally, a similarity score is estimated over the composed feature vectors. Experimental results show that our model gets the state-of-the-art performance on the answer sentence selection task, and achieves a comparable result on the paraphrase identification task.

In Proceedings of Coling 2016

References in corpus (1)

• Attentive Pooling Networks

Cited by in corpus (20)

• A Deep Network Model for Paraphrase Detection in Short Text Messages
• Bilateral Multi-Perspective Matching for Natural Language Sentences
• Multi-Perspective Context Matching for Machine Comprehension
• A Multi-cascaded Model with Data Augmentation for Enhanced Paraphrase Detection in Short Texts
• Multi-Cast Attention Networks for Retrieval-based Question Answering and Response Prediction
• $k$-Nearest Neighbor Augmented Neural Networks for Text Classification
• Neural Paraphrase Identification of Questions with Noisy Pretraining
• The Inception Team at NSURL-2019 Task 8: Semantic Question Similarity in Arabic
• LadRa-Net: Locally-Aware Dynamic Re-read Attention Net for Sentence Semantic Matching
• Improving Multilingual Semantic Textual Similarity with Shared Sentence Encoder for Low-resource Languages
• Sanity Check: A Strong Alignment and Information Retrieval Baseline for Question Answering
• M${}^2$S-Net: Multi-Modal Similarity Metric Learning based Deep Convolutional Network for Answer Selection
• Applying Transfer Learning for Improving Domain-Specific Search Experience Using Query to Question Similarity
• Learning Context-Sensitive Convolutional Filters for Text Processing
• A novel hybrid methodology of measuring sentence similarity
• BanditRank: Learning to Rank Using Contextual Bandits
• Inducing Alignment Structure with Gated Graph Attention Networks for Sentence Matching
• An Efficient Approach for Geo-Multimedia Cross-Modal Retrieval
• Learning review representations from user and product level information for spam detection
• Attentive Recurrent Tensor Model for Community Question Answering