Research And Implementation Of Web Service Discovery Based On Semantic Web

As a major component of the Service-Oriented Architecture, Web service discovery is aimed at searching appropriate candidate services for reusing, service composition and matching detection. Driven by Semantic Web technology, semantic description of Web services enhances the machine’s understanding ability of Web services, thus improving the accuracy of the Web service discovery. On the one hand, most of Web service providers do not provide semantic description for those services, and lower search performance of web service matchmaking based on ontology reasoning is not applicable. On the other hand, the current Web service description models are based on one-sidedness, which reduces the accuracy of Web service matchmaking algorithm. Finally, with the rapid increase in the number of Web services, it is one of challenging and important issues to be resolved at this stage that how to locate all appropriate candidate services for Web service requester among a larger number of Web services by an accurate and efficient way.In order to solve the key issues in Web service discovery, this paper maked a detailed analysis of Web service description and built Web service description model with text statistics and tree structural information. Based on this service description model, this paper uses tree homeomorphism algorithm and semantic matchmaking to create semantic Web service matchmaking algorithm, which solve the lack of semantic description of Web service and improve the shortcoming and deficiencies of those based on ontology reasoning. For the discovery problem in a large number of Web services, this paper used K-Means clustering algorithm to categorize registered Web services by function. This approach filters out irrelevant Web services by clustering information and enhances efficiency of Web service discovery. Under the absence of QoS information of Web service, this paper also proposed web service reputation evaluation model to select the best candidate Web services based on requester’s feedback. Finally, this paper designed and implemented clustering Web service discovery system based on semantic Web, which improve the efficiency and performance of Web service discovery.