The Research and Implementation of the Technology of Composition-Oriented Discovery of Semantic Web Service
|School||South China University of Technology|
|Keywords||service discovery service composition semantic web network diagram single-service-implementation|
SOC and SOA are milestones of technological progress in the fields of distributed systems and software integration. While, service discovery and service composition have problems of involving narrow and not completely satisfying needs. To solve these problems, composition-oriented service discovery arises. As the bridge between service discovery and composition, it contains both proceeding service discovery to provide service section discovered to service compositors, who composite these services efficiently for users to use them simpler. This paper studies composition-oriented service discovery and its approach. There are two main ideas: 1) Discovery - combined 2) decomposition - Discovery - combined. This paper study the existing deficiencies of composition-oriented service discovery approach based on hypergraph ideas. The specific content and research results are as follows:1) The concept of service single implementation is proposed. And then extract the single implementation of the service in the registry as available service data set. Available service data set together with atom services in the register center consist of basic service set which offers data set for semantic-oriented web service discovery. Also, analyzing users’requests, that leads to request sets and condition sets. 2) the study proposes and implements a network graph of the composition-oriented service discovery. Regarding service satisfying function needs as a node of the network graph, we build the network graph meeting the needs of requests which demand candidate services; 3) The I/O fuzzy matching which is based on ontology concept tree, provides basis for the similarity of I/O concept, measure of the relationship between service vertex and selection of service vertex; 4) The study proposes the concepts of candidate service access nodes and edges through its dependence set to reduce the access times of nodes in the process of network diagram analysis and achieving candidate service chains, while improving the efficiency of them.Finally, by comparing experimental and theoretical analysis, the method presented in the paper is proved to be valid and available.