The Study And Implementation Of Semantic Matching Algorithms In Semantic Web Serivces

As a new Web application model and remote access standard, Web Service is a perfect solution to integrating and sharing the data and application form heterogeneous platforms with the rapid development of Internet technology and applications. Because of the increase of Web Services, however, it makes more difficult to locate the services accurately and efficiently that match client's requirements. Traditional Web Services discovery mechanism is mainly based on key words comparing. The lack of machine-understandable semantic information leads to low recall ratio and low precision. Intelligent search can't be achieved. Thereby how to provide a more effective and efficient matchmaking mechanism is becoming a very important issue in automatic Web Services discovery based on semantic description of Web Services after introducing machine-understandable semantic information.First, this paper presents two similarity measures algorithms by considering all the determination such as ontology relationship and the depth of a concept. One is between two ontology concepts. And the other is between two concepts sets. The similatity which is used to describe how much the candidate service match the client's request will be obtained accurately and efficiently in use of the similarity measure algorithms. With meeting the functional requirements, a user preference matching algorithm based on the user interest model is added in. Then, all the services concerned will be output after being sorted in accordance with the integrated similarity. The filtering mechanism is introduced into the process of matchmaking in order to reduce the amount of similarity computing of the irrelative Web Services, which greatly improves the efficiency of the matchmaking algorithm.At last, this paper proposes a solution to facilitate the design and realization of Semantic Web Services matchmaking system. Travelling domain ontology and taxonomy ontology are built up to test the performance of the matchmaking mechanism. The test result verifies that this system based on similarity sorting and stages filter can improve precision and reduce the response time of Semantic Web Services matchmaking system.