Assessment of text coherence based on the cohesion estimation

TL;DR

This paper introduces a graph-based method for assessing text coherence through cohesion estimation, applicable across multiple languages, and demonstrates its effectiveness on English, Chinese, and Arabic texts.

Contribution

It presents a novel, language-independent graph-based approach for coherence evaluation based on cohesion estimation, expanding applicability to diverse languages.

Findings

Effective coherence assessment demonstrated on English, Chinese, and Arabic texts.

Method provides transparent understanding of the evaluation process.

Applicable to multiple languages without language-specific adjustments.

Abstract

In this paper, a graph-based coherence estimation method based on the cohesion estimation is suggested. Our method uses a graph-based approach to provide a user with an understanding of the evaluation process. Moreover, it can be applied to different languages, therefore, the effectiveness of this method is examined on the set of English, Chinese, and Arabic texts.