A Study of Ontology Engineering in Malay Unstructured Document Using Entity Relationship Model

Abstract

The existent of semantic web has encouraging the study in information retrieval and ontology based knowledge representation. Malay-language ontology engineering is not very encouraging due to Malay language technologies has limitation in Natural Language Processing and further more classified as an under-resource language. Therefore, appropriate technique is needed to be applied in Malay-ontology engineering. Thus, the objective of this paper is to find the best approach in generating Malay-ontology automatically. In this study we have develop MyOntologyGen prototype to create automatically Malay-ontology using natural language processing method and use entity relationship model as a conceptual model to develop in MySQL database. The evaluation strategy is to test the accuracy of ontology component against gold standard and model evaluation metrics. The result for average precision in taxonomy relationship between concept is 79% and non-taxonomy relationship is 63%. The analysis has shown that ontology component detection depends on the accuracy of group of word detection.