Research On Semantic-based Web Services Automatic Composition

Web services, because of its excellent heterogeneity, and solving the interaction data from different systems and reorganization, since the advent of the beginning, it has been welcomed by academics and industry and become a hot issue that people study. However, the service function published by provider the service function is specific, so a single Web service can not meet customer business needs of the complexity and variability, therefore, how to combine a large number of services of the different areas and different entities with a certain technology becomes particularly important. Manual and semi-automatic composition technology has now matured, whereas the automatic composition lack uniform norms and standards, people only put forward different composition programs and ideas from their respective areas. However, because of its good prospects and broad space for development, research on the automatic combination has been become increasingly enthusiastic. This research work is being done to meet this trend, Automatic Composition of Web services and related technologies have been made some work.First, introduces ontology and the critical technology of ontology building, discusses the latest W3C recommendation semantic annotation specification SAWSDL, on this basis, study documents of description of Web services, semantic annotation method, and then puts forward a matching algorithm based on matching degree from I/O interface matching, quality of service matching and function matching. Second, approaches of automatically building a Web services composition path based on input and output parameters associated with the rules mapping is put forward. This method firstly gives production rule bases, and models ontology according to input and output parameters, then from the user's inputs established mapping table which associated input and output parameters with rules and find out all combination of path meet conditions. Finally judging by quality of service the optimal schema is selected to return to users. And traditional service quality attributes and access mode and method of calculation has been improved, to make it more in line with the actual situation. We obtain QoS both from the service provider's initial value and consumer feedback data and continuously adjust and correct, and give full consideration to consumer concern about a particular attribute value when calculating the overall quality attribute value. The selection of the optimal combination of service programs to some extent depends on the consumer's preference for attributes. Finally, this paper describes shortcomings of the work and pointes out the direction of further study.This paper studies automatic Web service composition based on semantic, involving semantic annotation, matching and automatic composition, and explore the optimal combination of quality-driven program selection problem. Automatic Composition of Web services and related theory is full of challenges and inspires people to make more in-depth study.