Research On Web Service Discovery And Selection In Peer-to-Peer Environment

In today, web services have become the mainstream technology in Internet. Web services offer the distributed computing model of service oriented, it uses the simple and flexible agreement, the semantic definition and data expression of remote access service all use XML format which is the most popular, this message format supports the HTTP message binding, that solves the problem that heterogeneous system between service and client and the difference description language. Traditional web services make use of UDDI to complete the service discovery and selection, and do all operation in UDDI. In this situation, on the one hand this increase the burden of registration center server, on the other hand the web services system has Bottleneck problem. When the registration center server can not work well the whole web services system also can not work. Meanwhile, with an increase of number and kind of service, the efficiency of the service discovery and selection decrease.Based on the above issues, this paper research on the service discovery and selection in peer-to-peer environment, so we can make the service resources distributed storage, this solves the centralized system problem and improve the efficiency of the service discovery and selection. this paper construct two peer-to-peer that is a two layered peer-to-peer model according to the service discovery and selection, and put the similarity of ontology in the service discovery. The three layered matching which is ontology matching, based on keyword matching and semantic matching between service advertisements and service requests by logical reasoning is proposed in this paper. We select the first two matching as the criteria for domain and group classification in construction of peer-to-peer.In service selection, choosing QoS as the criteria for service selection. The selection method is that we select the service whose QoS is closest to the requestor's QoS instead of whose QoS is the best. This method solve the load problem of Web services provider. This paper apply this method in peer-to-peer and choose the similarity of QoS as the criteria for group classification in construction of peer-to-peer based on service selection.This paper combine Web services technology with peer-to-peer technology in order to complete the service resources distributed storage and avoid the bottleneck problem. Experimental results show that given an appropriate setting. this can improve the efficiency of the service discovery and selection...