Thai-sentence classification using conceptual graph

Abstract

An online discussion plays an important role in collaborative learning (CL). It can be used to track the behavior and cooperation among students and exhibit them in terms of interaction patterns. The essential mechanism used to describe these interaction patterns is a sentence classification. Although many papers have been reported sentence classifications in various languages, no research papers introduced Thai-sentence classifications. Therefore, this paper proposes an automated Thai-sentence classification using a conceptual graph (CG). In this method, there are two cooperation sectors: a conceptual graph and a semantic analyzer. The former represents a knowledge base of a Thai grammatical structure, which is designed in an XML form. It is called an XML conceptual graph. The latter is a semantic analyzer identifying types of sentences by matching a token with the XML conceptual graphs. The experimental result shows that the proposed system can accurately classify Thai sentences in online discussions.