Thai Rhetorical Structure Analysis

TL;DR

This paper presents a novel approach for Thai rhetorical structure analysis, utilizing specialized segmentation, clustering, and decision trees to identify discourse relations in Thai texts, aiding various NLP tasks.

Contribution

It introduces a unique method tailored for Thai language, combining HMM-based EDU segmentation, semantic-rule-based clustering, and decision trees for discourse relation classification.

Findings

Effective segmentation of EDUs in Thai using HMMs

Successful construction of rhetorical structure trees based on semantic similarity

Accurate classification of discourse relations with decision trees

Abstract

Rhetorical structure analysis (RSA) explores discourse relations among elementary discourse units (EDUs) in a text. It is very useful in many text processing tasks employing relationships among EDUs such as text understanding, summarization, and question-answering. Thai language with its distinctive linguistic characteristics requires a unique technique. This article proposes an approach for Thai rhetorical structure analysis. First, EDUs are segmented by two hidden Markov models derived from syntactic rules. A rhetorical structure tree is constructed from a clustering technique with its similarity measure derived from Thai semantic rules. Then, a decision tree whose features derived from the semantic rules is used to determine discourse relations.