It is very important to find suitable services for information sharing and reuse in the dynamic, heterogeneous network environment.Both the industrial sector and academic sector pay close attention to web service discovery.Because traditional technical standards of web service is lack of the support of logical semantics,there are still many problems to issue about the research of web service discovery.So,this paper studys semantic web service discovery.Firstly,the ways of semantic description of web service and the WSDL/Owl-s mapping mechanism are studied.Secondly,in order to improve the traditional UDDI-based service discovery architecture,a new framework based on the expansion of Augment UDDI registration system for web service discovery is designed.As a result, the new framework makes effective support for the search and registration based on semantic.Thirdly, for the coarse granularity of the matching results and the negligence of the potential demand of service quality of classical elastic matching algorithm,Owl-s is expanded,and a multi-level matching strategy based on ontology concept semantic similar degree for web service matching is designed,which involves the function matching,the category matching,the text matching and the quality matching. So,a better matching granularity is achieved,and the service matching average recall and precision are improved,too.Finally,the key modules of the web service discovery framework are implemented.To verify the validity of the framework,the query of web service of experiment based on the ontology of computer books online is maken.The results show that the performance of the framework is greatly improved. |