Research On B2B Integration Based On Semantic Web Services

Traditional B2B integration technologies are not suitable for the heterogeneous Internet environment. The Web Service technology is widely used currently, however, for its lack of semantic information, the computer can not really"understand"the meaning of resources on Web. So the heterogeneous business process and too much manual intervention problems in B2B integration can't be well resolved. Aiming to resolve above problems, this paper combined Web Service technology with Semantic Web technology. By taking advantages of platform-independent, loosely coupled and other characteristics of Web Service, and through the semantic description and annotation on various resources in B2B integration process, the Web services are semantic. The mediators in WSMO are well used to resolve the data and process heterogeneous in B2B integration process. The service matching, discovery, composition and invoking can be automatic so as to realize the automation of B2B e-commerce procurement process. The main contributions and innovation of this paper are listed below:(1) A B2B integration framework was modeled based on semantic Web services. The problems urgent to be solved in B2B integration are the heterogeneous business process integration problems. Compared with other semantic Web services frameworks, the biggest advantage of WSMO is providing mediator to resolve heterogeneous problem. So we choose WSMO as our integration model and established B2B integration framework based on WSMO. We also discussed the whole process of service description, discovery, composition and invoking under this framework.(2) A semantic Web service composition algorithm was proposed. This paper analyzed the drawbacks of current Web service composition studies—the semantic and quality of services (QoS) were not considered comprehensively. On this basis, a service composition algorithm based on Bipartite Graph optimal Matching and QoS was proposed. Considering the service semantic information and QoS together, we made use of bipartite graph optimal matching algorithm to calculate the degree of semantic association. And we used multi-objective decision-making methods to measure the service quality satisfaction. The test simulation proved that AL-CWS can automatically realize Web service composition according to service request, and can assure the quality and efficiency of service composition by controlling threshold value.(3) On the basis of above research, a prototype system was proposed which is an application of this research combined with practice. And it also provided a better solution for the application of semantic Web services in B2B integration.