Thai Succession and Family Law Ontology Building Using Ant Colony Algorithm

摘要

Ontology building is a tedious job and a time consuming task for user. The quality of ontology plays an important role in information retrieval application. Therefore, we need the expertise from law users to extract their knowledge in term of ontology structure. To overcome these difficulties, we develop an algorithm using ant colony to automatically extends a seed ontology that was initially created by law expert. Two seed ontologies are constructed by experts which are succession law and family law ontology. These seed ontologies can be extended using Supreme Court sentences as a corpus. We used 2 datasets from civil Supreme Court sentences, which are succession law and family law. Our ontologies are embedded with weight values that are the product of pheromone updating process of ant colony. We investigate the experimental results and found that the weighted ontology provides significant improvement on the retrieval process. Furthermore, the size of our ontologies obtained from TLOE is suitable since their retrieval performances are acceptable by users.