An Analysis of Sentential Neighbors in Implicit Discourse Relation Prediction

TL;DR

This paper investigates the impact of broader contextual information on implicit discourse relation prediction, finding that including context beyond immediate neighbors can be detrimental to classification accuracy.

Contribution

It introduces three novel methods for incorporating context in discourse relation prediction and evaluates their effects on classification performance.

Findings

Including broader context beyond immediate neighbors is harmful.

Proposed methods offer new ways to incorporate context.

Contextual expansion does not improve classification accuracy.

Abstract

Discourse relation classification is an especially difficult task without explicit context markers (Prasad et al., 2008). Current approaches to implicit relation prediction solely rely on two neighboring sentences being targeted, ignoring the broader context of their surrounding environments (Atwell et al., 2021). In this research, we propose three new methods in which to incorporate context in the task of sentence relation prediction: (1) Direct Neighbors (DNs), (2) Expanded Window Neighbors (EWNs), and (3) Part-Smart Random Neighbors (PSRNs). Our findings indicate that the inclusion of context beyond one discourse unit is harmful in the task of discourse relation classification.