Research On Domain Ontology Semi-automatic Construction In Semantic Web Services

At present,the method to realize Semantic Web Services is to use ontology to describe Web services to enable the automated discovery,composition and invocation of Web services. But, the existing ontology construction process has a series of problem, such as:the process complex,too rely on experts,the construction speed slow and terminology expression imprecise. As an effective solution to solve the above problems,the research on semi-automatic or automatic ontology construction,reducing the dependence of experts and speeding up the speed of ontology construction become more and more important.After studying existing domain ontology construction methods,this paper puts forward a domain ontology semi-automatic construction framework which is to construct an ontology with the thesaurus and reuse ontologies in ontology base and Web to perfect this ontology.This paper discusses about the basic theoretical knowledge of ontology and thesaurus,gives the rules for the convertion from thesaurus to ontology,successfully converts the thesaurus to an ontology using the method of describing ontology with OWL.While perfecting the init ontology,according to the characteristics of thesaurus ontology,this paper changes the existing method of calculating semantic similarity and designs an integrated semantic similarity calculation formula,the formula can be more comprehensive and accurate calculating the semantic similarity between concepts.While perfecting these ontologies,this paper gives the corresponding rules,processes and algorithms of adding data,and improves the automation of ontology construction.At last,this paper designs and realizes a prototype system which can realize domain ontology semi-automatic construction in Semantic Web Services,and can solve the problems of the process complex,too rely on experts,the construction speed slow and terminology expression imprecise.The experimental results reveal that the framework and algorithm are feasible and effective.The research result of this paper not only speed up the process,but also improve the ontology rationality,normalization and authority,and that would also be a good way to build domain ontology.