Neural Discourse Structure for Text Categorization

Abstract

We show that discourse structure, as defined by Rhetorical Structure Theory and provided by an existing discourse parser, benefits text categorization. Our approach uses a recursive neural network and a newly proposed attention mechanism to compute a representation of the text that focuses on salient content, from the perspective of both RST and the task. Experiments consider variants of the approach and illustrate its strengths and weaknesses.

ACL 2017 camera ready version

References in corpus (3)

• A Convolutional Neural Network for Modelling Sentences
• DyNet: The Dynamic Neural Network Toolkit
• Efficient Character-level Document Classification by Combining Convolution and Recurrent Layers

Cited by in corpus (6)

• Learning Structured Text Representations
• Causal Explanation Analysis on Social Media
• GILE: A Generalized Input-Label Embedding for Text Classification
• Improve Document Embedding for Text Categorization Through Deep Siamese Neural Network
• Evaluating Document Coherence Modelling
• Corruption Is Not All Bad: Incorporating Discourse Structure into Pre-training via Corruption for Essay Scoring