RST Parsing from Scratch

TL;DR

This paper presents a novel end-to-end RST discourse parser that models parsing as token boundary splitting decisions using a seq2seq network, enabling segmentation and parsing without handcrafted features.

Contribution

It introduces a top-down, sequence-to-sequence approach for discourse parsing that performs segmentation and tree construction simultaneously, without relying on prior segmentation.

Findings

Outperforms existing methods in end-to-end RST parsing

Does not require handcrafted features, improving speed and adaptability

Achieves better accuracy on the standard English RST discourse treebank

Abstract

We introduce a novel top-down end-to-end formulation of document-level discourse parsing in the Rhetorical Structure Theory (RST) framework. In this formulation, we consider discourse parsing as a sequence of splitting decisions at token boundaries and use a seq2seq network to model the splitting decisions. Our framework facilitates discourse parsing from scratch without requiring discourse segmentation as a prerequisite; rather, it yields segmentation as part of the parsing process. Our unified parsing model adopts a beam search to decode the best tree structure by searching through a space of high-scoring trees. With extensive experiments on the standard English RST discourse treebank, we demonstrate that our parser outperforms existing methods by a good margin in both end-to-end parsing and parsing with gold segmentation. More importantly, it does so without using any handcrafted features, making it faster and easily adaptable to new languages and domains.