Study on Web Service Matching and Composition Based on Ontology

Abstract

Because web service function is continually strengthened, and web service quantity increases rapidly, it is urgent to solve the problem on automatic service matching and composition to meet service request. The service matching algorithm for bionic manufacturing domain-specific ontology is proposed in the paper. For this algorithm, the semantic similarity between two concepts is calculated according to the distance of two nodes corresponding to two words of service request and description. The revised index for matching the key word is considered to emphasize the key parameter among service description, and the matching accuracy is highly improved. The total matching degree of service is equal to the average of all semantic similarities between the service request and the service description. It can be accurately judged if the service matching is successful by the total matching degree. According to the dynamic hierarchical structure of domain-specific ontology model of bionic manufacturing, the service auto-link composition algorithm is proposed based on the recursive principle in the paper. The service is automatically and intelligently matched and reasoned by using the heuristic knowledge of knowledge base and the semantic information of ontology base of bionic manufacturing. The structure-well user objective IOPEs (Input, Output, Precondition, and Effect) is linked to get the abstract or concrete service-flow framework. The automatic service composition is implemented from single service to higher concept hierarchical service. Lastly, the system structure of semantic web service matching and composition is designed based on two described algorithms