Research On Semantic Web Services Composition

Web services technology is rapidly emerging in a large variety of new paradigms such as business-to-business(B2B) interaction,process integration,deployment, and management.The reason for this great success achieved in Web services arena can be boiled down to its declarative lookup and invocation modes.In particular, Web services highly rely on some descriptive XML-based artifacts to accomplish communication and interaction,henceforth some de-facto standards,such as SOAP, WSDL,and UDDI,are speedily shaping to accelerate the development of Web services agenda.Unfortunately,those existing standards seems to be incapable of bearing such a burden to describe the functional and non-functional requirements in a machine-understandable way and thereby achieve a fully automated verification process.As the complement,Semantic Web notion targets at semantically marking Web-accessible resources,and this leads to a natural expectation to combine Semantic Web and Web services together.In this spirit,the so-called Semantic Web Services(SWS) notion is identified from those two trends,but a number of obstacles still remain unsolved nonetheless.In this paper,according to different stages of Web services composition lifecycle,we address our work on four aspects:process modeling,semantic verification,rule analysis,and exception handling.First of all,we elaborate a formal variant of Description Logics,namely, ALCQO(Q~*),which is capable of extending the conceptual modeling ability of classical DLs with process modeling and numerical constraint representation ability, while still retains decidability for major reasoning problems.Secondly,we propose a general modeling and verification mechanism for Web services composition.To provide a general guidance for users,we articulate a suite of transformation rules to automatically rewriting a workflow-based Web services contracting process into our ALCQO(Q~*) framework,while can be regarded as harmless w.r.t,the overall reasoning complexity.Moreover,the verification process can be built on the transformation result and benefit from its soundness and completeness.Thirdly,we propose a semantic analysis mechanism for business rules.We choose a variant of Description Logics(DLs),called ALCQO(Q~*),as the underlying logic,and provide a formal mapping to transform ECA rules,so that the semantics in the original ECA rules can be captured and are computationally traceable. To this end,we further investigate some important properties for business rules,namely,redundancy,termination and conflict,and propose several sound and complete algorithms to resolve them.Lastly,we propose to capture exception contexts to facilitate exception handling in a composite service.Particularly,we present a mechanism to capture external exception contexts from partners' applications through Contexts enhanced Web Services Conversation(CeWSC),which introduces just a minimal overhead by an effective use of SOAP header and acknowledgement.Moreover,we also propose an enhanced event-driven mechanism to incorporate exception contexts into exception handling for a composite service to facilitate its flexible and reliable execution.