Testing SDRT's Right Frontier

TL;DR

This paper empirically validates SDRT's Right Frontier Constraint, showing it is respected in most discourse structures and highlighting its significance for discourse interpretation and machine learning applications.

Contribution

It provides strong empirical evidence supporting SDRT's version of the Right Frontier Constraint through analysis of annotated discourse documents.

Findings

SDRT's RFC respected about 95% of the time in analyzed documents

Violations are mostly due to annotation errors or misconceptions

Supports SDRT's applicability in discourse analysis and ML

Abstract

The Right Frontier Constraint (RFC), as a constraint on the attachment of new constituents to an existing discourse structure, has important implications for the interpretation of anaphoric elements in discourse and for Machine Learning (ML) approaches to learning discourse structures. In this paper we provide strong empirical support for SDRT's version of RFC. The analysis of about 100 doubly annotated documents by five different naive annotators shows that SDRT's RFC is respected about 95% of the time. The qualitative analysis of presumed violations that we have performed shows that they are either click-errors or structural misconceptions.