Two Ontological Approaches to Building an Intergrated Semantic Network for Yami ka-Verbs

This paper describes a proposed ontological language processing system for integrating two semantic sets for a group of important verbs with the prefix Kain Yami, an Austronesian language in Taiwan. The two semantic sets represent two different classification approaches. One approach follows the concepts and rules of WordNet and the other uses the metaphors in Yami indigenous knowledge. The ontologies are used for classification and semantic integration. The results of implementation are used for building the Yami lexical database. This paper illustrates how the methodology and framework used in classifying Yami can be applied to Austronesia language processing.