Reserch And Implementation On Semi-Automatic Domain Ontology Acquisition Method

Ontology is a kind of model used to describe the concepts and their relationships. Since proposed, it has attracted a lot of researchers in the world to focus on this field and has been applied widely in the computer science. Ontology building method, as a base work in the ontology research field, has also attracted many researchers. The available ontology building environment can meet the need in the construction of ontology now, but human experts have to make huge efforts for constructing them through finding concepts and their relationships by hand. So the application based on ontology can hardly been used abroad.In this paper we research on semi-automatic domain ontology acquisition by applying statistical natural language process and text mining technology to the process of ontology building. The domain relevant concepts and their relationships can be acquired from domain corpus and then human editor verify the results of the acquisition. When human editor build ontology, he or she can combine the results got from automatic step to the ontology built by hand, so ontology building process is more rapid than before.The thesis is structured as follows: Chapter 1 is the introduction. In chapter 2, we introduce the relevant descriptions of ontology and the process of ontology construction. In chapter 3, we discuss how to acquire domain concepts through mining domain corpus. In chapter 4, we discuss how to mining the relationships of concepts from domain corpus. In chapter 5, we introduce the realization of prototype system and test the algorithms adopted in the system by applying to a really specific domain. The last chapter is summarization and expectation.