Service Composition Discovery Oriented To Web Service Set

With the development of social science and technology,the information network has been developed rapidly.The social efficiency and the quality of life can be improved by information network.Web service is the software interface based on information network.Web services in remote side can be used in local computers,such as weather report,network payment,tickets booking,and so on.Web services are defined based on XML.The SOAP protocal is used for communication between remote Web service side and the local computer.The software platform can be constructed based on Web services.The efficiency and integration to construct the software platform can be improved.With the popularization of the Web service,the number of Web service is increasing.Some data showed that there are tens of millions of Web service in the word.Different interfaces and the different functions can be supplied by the different Web services.Although there are more Web services,the efficiency and precision of service composition discovery is decreased.In order to sovle the above problems,the methods for service composition discovery oriented to Web service set are proposed in this paper.First of all,the formal definition of Web service with Qos characteristic is given.The technology of service clustering is introduced.Secondly,the logic Petri nets are enhanced according to the definition of Web service.The analysis of reachability of enhanced logic Petri net is showed.The theorem of judging enabled transitions is given.The algorithms of marking computing after a transition fired and obtaining the reachable marking are proposed.Thirdly,the logic Petri net model of Web service set is given.The strategy of service composition discovery with the global parameter constraint is introduced.Moreover,the method for service composition replacement is given.At last,the methods proposed in this paper are proved to have superiority in improving the efficiency and precision of service composition discovery by the simulation experiment.