RST-style Discourse Parsing Guided by Document-level Content Structures

TL;DR

This paper introduces a new RST discourse parsing pipeline that leverages document-level content structures from news profiling to improve the prediction of discourse relations in large text spans.

Contribution

It proposes a novel RST-DP pipeline that integrates high-level content representations, enhancing parsing accuracy for large discourse spans.

Findings

Improved RST parsing performance across multiple metrics

Effective use of content-aware sentence representations

Minimal additional computational layers required

Abstract

Rhetorical Structure Theory based Discourse Parsing (RST-DP) explores how clauses, sentences, and large text spans compose a whole discourse and presents the rhetorical structure as a hierarchical tree. Existing RST parsing pipelines construct rhetorical structures without the knowledge of document-level content structures, which causes relatively low performance when predicting the discourse relations for large text spans. Recognizing the value of high-level content-related information in facilitating discourse relation recognition, we propose a novel pipeline for RST-DP that incorporates structure-aware news content sentence representations derived from the task of News Discourse Profiling. By incorporating only a few additional layers, this enhanced pipeline exhibits promising performance across various RST parsing metrics.