An Automatic Method of Finding Topic Boundaries

TL;DR

This paper introduces an automatic method for identifying discourse boundaries using lexical cohesion and dotplotting, with experiments demonstrating its effectiveness in segmenting concatenated documents.

Contribution

It presents a novel combination of lexical cohesion analysis and graphical dotplotting for automatic discourse boundary detection, including an optimization algorithm.

Findings

Successful automatic boundary detection in concatenated texts

Effective use of dotplotting for discourse segmentation

Potential applications in text analysis and processing

Abstract

This article outlines a new method of locating discourse boundaries based on lexical cohesion and a graphical technique called dotplotting. The application of dotplotting to discourse segmentation can be performed either manually, by examining a graph, or automatically, using an optimization algorithm. The results of two experiments involving automatically locating boundaries between a series of concatenated documents are presented. Areas of application and future directions for this work are also outlined.