A Research For Semantic Web Service Discovery Based On P2P

With the widely application and promotion of web services, the web service technology has already become an effective resolution for service-oriented computing of the next generation. There’re so many web services provided on the network with similar functions, but their non-functional properties vary greatly. It’s difficult to discover services that could both meet user’s functional and non-functional requirements from huge amount of web services. So how to build a useful service discovery mechanism has been a host spot and key issue in the research field of web services.Referring to these existing research works, this paper starts its study with semantic web services discovery mechanism in the peer-to-peer environment. Firstly, it improves the balanced nested tree, and extends the tree with semantic information, based on which a semantic nested balanced distribution tree model is proposed. A series of web services operations have been implemented in this model, such as service discovering, new service adding and invalid service deleting. Besides, this paper has further developed a new services discovery mechanism which integrates the functional semantic with quality property of services. The new mechanism filters the whole services according to domain experts and users’ personal preferences, then normalizes the quality parameters of the remaining services. In order to find those services that satisfy the users most, it weights the functional semantic similarity and QoS attributes which users are interested in at last.The services discovery simulation result has shown that this service discovery mechanism takes more time than other methods in the model building period, but provides with much better query efficiency and recall ratio. Thus it has great advantage under the frequently query scenarios. In addition, this paper verifies the rationality and effectiveness of the new mechanism by a ticket booking example.