Researching The Model Of Semantic Web Service Composition Based On The Petrinet

With the mushroom of the network technology and its application, people's requirements to network are enhanced. Thus, various services which were able to satisfy all requirements effectively were expected. The center of gravity for computer network research has been extended to the application layer in network. The next generation network in the future would be a user-oriented, QoS-assured network which focuses on service. It would provide a lot of various services and has the function to customize and manage these services. To complement discovery,call,composition and excuting monitor for the automatic web service, semantic web service technology uses ontology description language to make comprehensible semantic label to web service decription.Auto discover and select service is key of composing semantic web service automaticly, by contrast, the more important thing to semantic web service is the analysis of semantic web service composition such as reachability,safety and detect the possible deadlock. Currently, the semantic service composition exsists two problems as follows: 1. lack validation mechanism. WSDL OWL-S or BPEL4WS do not provide validation mechanism to their web service description (include ontology) totally.2.lack integrase application environment for developing web service.Validation mechanism was added into web service composition, which provides an exact service composition plan for users'requirement and without overmany artificial manipulation. After making deep analysis and study for web service composition technologies and formal validation of Petri net, this paper adopted OWL-S to describe web service, proposed the semantic web service composition based on Petri net. Petri net description of semantic web service composition could not only describe the various details in process of service description clearly and exactly, but also simulate and dynamic analysis to the executive instance of semantic web service, which included qualitative and quantitative analysis.Currently, the tools which related to web service description,validation and composition are various and emphasis particularly,such as Jena,OWL-S Editor. They could be used to generate semantic web service description,but lacking validation and implement function of service composition. Besides, the popular validation tools always lack service description and discovery mechanism,such as JCPN,VDM. The BPEL4WS execution engine could execute web service described by BPEL and WSDL, but could not execute web service described by semantic description and execution sequences described by Petri net. So this paper integrated several tools and environments, build a uniform application model.In process of web service match, there would appear composition glossary composed by several concepts inevitablely, which not only lack definition but also composition mode and structure have randomicity in current semantic dictionary such as WordNet. The order among glossary lacked uniform definition. The composition among glossary did not have uniform rule, and sense of composition correlated to each glossary.therefore, this paper proposed a new method to calculate similarity of several glossary composition.