A Construction of Thai WordNet through Translation Equivalence

Abstract

WordNet is a crucial language resource associated with artificial intelligence activities, for instance, constructing building models for advancement of computational linguistics and natural language processing, or representing statistical insights through knowledge graphs that emulate cognition and human understanding. Thai WordNet has been developed in many approaches, e.g., a merge approach in gold standard, and semi-auto construction with a bilingual dictionary. However, existing Thai WordNet is not easy to find words fit with the definition of synsets; and cover cultural gaps between the different languages of which needed to be aware. This paper presents a methodology of Translation Equivalence in order to construct Thai language resource, called LST22 Thai WordNet.