Research And Implementation Of Semantic-Based Web Service Hybrid Matchmaking Strategy

With the global wave of information, information technology industry continues to develop, extend, has deep into many companies and individuals, the emergence of SOA system architecture, bring a new revolution for the enterprise information. SOA’s feature of loosely coupled make it become the foundation of modern enterprise system. The Web service technology as one of the key supporting technology of SOA, has already obtained more and more widespread attention.For the past few years, research works in the area of web service focus on various issues arise throughout web service life cycle. Include how to specify、 discover, select, mediate, compose,invoke and monitor web services. Because of the inconsistency and incompleteness of web services, it is very difficult for the computer to accurately find the web service meet the needs of users. Web service discovery has become one of the key problems which restrict the application and development of Web services.With the semantic web is put forward, combining it with the Web service, reasonable use of the semantic information of Web services, can effectively support the discovery of Web services technology automated processing, which can realize the seamless integration and interoperability between heterogeneous systems. Semantic web services contains rich semantic information in the service description, Service discovery algorithm can use these semantic information for service matching, So as to overcome the shortcomings of the service discovery mechanism based on keywords.In this paper, we discusses and analyzes the current Web service matching strategy based on semantic, and propose an improved hybrid matching method of Semantic Web services. The matching method includes two stages. Firstly, operation-based matching(mainly the input and output parameters matching).the bipartite graph matching method is introduced, and in this phase bipartite graph matching method is improved, and the semantic similarity of ontology concepts as the weights of bipartite graph for the similarity computing Secondly, TextDescription-based matching.the short-text matching method is introduced, and extracting the grammatical structure of the text, and computing concept similarity with the same grammatical role, and then integrated similarity of different grammatical roles, finally obtaining the whole text similarity.Finally, based on the service discovery method proposed in this paper set up the experimental platform, and the OWLS-TC as the test set,through the compared with OWLS-MX2. the experimental results show that the proposed matching strategy can effectively improve the efficiency of services discovery.