Domain Ontology Generation Based on WordNet and Internet

This paper presents a method of extracting domain ontology based on WordNet and Internet. Domain concepts are extracted from WordNet which contains concepts of almost all the fields. The extracting method is based on the IC-based semantic similarity, this paper gives a new model of semantic similarity which has better performance and is the guarantee of extracting domain concepts more accurately. In order to get the taxonomy relation between the domain concepts, the approach of taxonomy learning from Internet is given, which is based on the mutual information matrix. The generation method makes full use of the WordNet and Internet and can improve the efficiency of domain ontology generation.