Deconvolutional Latent-Variable Model for Text Sequence Matching

Abstract

A latent-variable model is introduced for text matching, inferring sentence representations by jointly optimizing generative and discriminative objectives. To alleviate typical optimization challenges in latent-variable models for text, we employ deconvolutional networks as the sequence decoder (generator), providing learned latent codes with more semantic information and better generalization. Our model, trained in an unsupervised manner, yields stronger empirical predictive performance than a decoder based on Long Short-Term Memory (LSTM), with less parameters and considerably faster training. Further, we apply it to text sequence-matching problems. The proposed model significantly outperforms several strong sentence-encoding baselines, especially in the semi-supervised setting.

Accepted by AAAI-2018

Cited by in corpus (10)

• How to Fine-Tune BERT for Text Classification?
• A Tutorial on Deep Latent Variable Models of Natural Language
• Baseline Needs More Love: On Simple Word-Embedding-Based Models and Associated Pooling Mechanisms
• Towards Generating Long and Coherent Text with Multi-Level Latent Variable Models
• Enhancing Dual-Encoders with Question and Answer Cross-Embeddings for Answer Retrieval
• Cross-modal Variational Auto-encoder for Content-based Micro-video Background Music Recommendation
• Enriching Conversation Context in Retrieval-based Chatbots
• NASH: Toward End-to-End Neural Architecture for Generative Semantic Hashing
• Crossing Variational Autoencoders for Answer Retrieval
• Semi-Supervised Dialogue Policy Learning via Stochastic Reward Estimation