Research And Implementation Of On Semi-automatic Ontology Construction Base On WordNet And Focused Crawler

Ontology is an explicit specification of a conceptualization. The term is borrowed from philosophy, where Ontology is a systematic account of existence. For AI systems, what exits is that which can be represented. Since 1990s, ontology has been introduced in AI, Knowledge Engineering and Library & Information sciences. It has been applied widely in the computer science. Ontology could solve two important knowledge engineering problems: knowledge sharing and knowledge reusing. Though manual construction is comparatively accurate, but it has many disadvantages such as complexity of the project, time-consuming, hard sledding and strongly depending on the domain expert's subjective consciousness. On the other side, with the development of information technology and new knowledge is proposed constantly. Therefore, the problem about ontology upgrade is important.In this paper we research on semi-automatic domain ontology acquisition by WordNet and focused crawler technology to the process of ontology construction. WordNet based as the kernel, focused crawler for auxiliary, calculating concept similarity, the domain relevant concepts is built. Then, the relationship of concepts is acquired from hierarchical clustering arithmetic based on single linkage. Finally, the human editors verify the results of the acquisition.Some achievement the paper has acquired can be expressed as follows:(1) Though introducing the relevant descriptions of ontology and the process of ontology construction, propose semi-automatic ontology construction base on WordNet and focused crawler.(2) Take WordNet as semantic dictionary, use its own semantic relations, calculate information content of concepts, and then obtain semantic similarity.(3) Focused crawler works as large-scale corpus, computes the frequency which appears in the context, and then calculates semantic similarity.(4) According to formerly calculated semantic similarity, use hierarchical clustering arithmetic based on single linkage, and then ontology integration.(5) Develop ontology tool—MyProtégé, users can modify ontology according by condition.