Match The Semantic Web Services Technology Research

With the development and applications of Service Oriented Architecture, there are more and more services which are supported by multiplatform and multilanguage, extensible, reliable and secure welling up on the web. Semantic Web was brought forward with the intention of matchmaking and finding the services automatically in time according to the request of the user. The existing web service matching is mainly based on key words comparing, however, the lack of semantic information leads to low recall and precision and can't implement intelligent finding.Thereby, to provide a more accurate and more efficient matchmaking mechanism based on semantic description of web services is becoming a very important issue in automatic web services discovery.To solve the above problem,the main research works I do as follows:(1) The knowledge about web service and semantic web is introduced allsided,including the latest progress. On the foundation of studies about the information field of semantic similarity algorithm,this paper proposes an improved single-concept ontology similarity measure.Combing the given semantic distance calculation algorithm as well as constructing a suitable similarity function,the similarity algorithm can efficiently and accurately calculate the similarity between two lotology concepts by using the proposed weight distribution method and the construction of the node route table to pre-process the relations of ontology's inheritance and property;Based on the single-concept ontology similarity measure, the information field of a classic concepts sets similarity algorithm is improved so that it can be applied to semantic web service matching. Simulation results demonstrate that the improved algorithm of matching accuracy outperforms traditional similarity measures.(2) This paper proposes a two-dimensional model of describing web services based on semantic, which describes the service's attribute about syntax and semantic informations after analyzing WSDL and OWL-S.The model supplies the WSDL's shortcoming of semantic and reinforces the semantic information of the OWL-S, which supplies good sematic foundation for the matchmaking of web services.On the foundation of the two similarity algorithms, a macthmaking model is designed in this paper according by the two-dimensional model of describing web services and the given service matching strategy.In order to reduce the similarity computaion of irrelevant service, the use of classification and matching filter mechanism can greatly improve the efficiency of web services matching on the foundation of combination with the proposed similarity algorithm.(3) This paper provides a solution to facilitate the design and realization of semantic web services matchmaking algorithm and verification system, at the same time, the matchmaking algorithm and the matchmaking model are verified on the basis of OWL-S Profile documents described by two-dimensional description model, the results desmonstrate that they are effective and feasible. They have very good effect in improving Web Services precision,recall and can meet the requester's requirements.