基于p2p的web服务聚类及组合研究-计算机应用技术专业论文.docxVIP

Title: RESEARCH ON THE CLUSTERING AND COMPOSITION OF P2P-BASED WEB SERVICES Major: Computer Application Technology Name: Xiaobo HU Supervisor: Prof. Lei WANG Signature: 泊加/，o/ftJ. Signatur区2场/ Abstract Along with the fast development of network technology and expanding network ，more and more web service can be obtained from the internet.But the cuηent widely used service discover technology which is centralized on key words and semantics leads to a bottleneck in registration center and strongly affects the efficiency of services discovery and matching.How to obtain web service rapidly，accurately and comprehensively becomes a hot research topic. For the problem during service discovery and matching，a novel model of web services is probed which makes use of some ideas from the peer-to-peer structure and the clustering methods. In this model ，it divides centralized registration center into P2P service communities located both logically and physically ，then cluster dynamically web services according to their semantic function in communities.The service discovery and matching will follow these two steps: Firstly，break up the users objective requirement into a series of smaller objective requirement ，search relevant cluster service based the semantic function of these smaller objective requirements. Secondly，match service exactly based on the input and output parameter of service requirement in clustering service. Per the above research ，the single web service can not meet peoples need in network ， while the value-added web service combination can. So this article further probes and studies a forward chaining web service combination model. This model can automatically combine and optimize discovered and matched web service，feedback the combined and value-added service to the users to satisfy their needs. The theory analysis and simulation experiment shows the model and algorithm mentioned in this article improves a lot in web service matching speed and pre