Web Service Composition via Semantic Matching of Interaction Specifications

Web service composition is a difficult and important requirement for the success of the Semantic Web. In this paper, we present a method of facilitating automated service composition through the declarative specification of potential interactions between them (or workflows) and the provision of semantic matching support to allow the automated interpretation of the interaction specifications. The key idea is that matching allows us, at run time, to find the correspondences among independently defined heterogeneous web service descriptions. Our approach extends existing methods to allow interaction models to convey relevant matching information and develops novel semantic matching techniques which allow for the interpretation of logical terms. The preliminary evaluation results show high efficiency and effectiveness of the proposed matching techniques.