Association Rule Based Context Modeling for Web Service Discovery

Web service is essential in achieving dynamic business process. With the dramatically increased number of Web services advertised in UDDI, Web Portal or Internet, how to locate the best Web service according to a user´s requirement is becoming more and more important, which calls for efficient and effective Web service discovery mechanism. Considerable efforts have been attained in solving this problem among them semantic based approaches show encouraging result. However, when several semantically equivalent Web service candidates are returned by matchmaking process, how to discern which one is the most suitable one is a real challenge. This paper presents a framework by which provider´s context and user´s context has been explored to help understand the real need of a user. This framework uses association rule for context modeling of provider and user from Web service composition perspective. After modeling context, a ranking service is invoked to compare Web service candidates with user´s requirement and then the best suitable potential Web service will be selected.