Research Semantic Web Services Matching Technology Based On Similarity

Requester of Web Services can find out satisfied service results through Web Services discovering technic. However, traditional Web Services discovering technic is done by keyword match based on the syntactic description of Web Services, so it results in the lower recall and precise and participation of requesters. In order to raise the recall and precise and become reality for making computer to finish automatically service discovering, it is necessary that add semantic information which machine can comprehend and process. Semantic Web Services discovering technic is becoming a hot research topic.This thesis introduces roundly Web Services and its discovery mechanism and analyses two key problems in the semantic Web Services discovering technology----Services description and Services matching. A food domain ontology is constructed in this thesis. Some concepts in that ontology are used to annotate the semantic of Web Service, the problem of finding the fittest Web Services is transformed to the calculation of the semantic distance and semantic similarity between two concepts in the domain ontology.Based on accurate description of concepts and relations among them in domain ontology, this thesis makes some research on present semantic distance and improves semantic distance algorithm, proposes algorithm of semantic distance for Corss-Ontologys.The present semantic similarity algorithms are analyzed and a hybrid similarity measure is proposed based on semantic distance and dynamic weight.At last, this thesis designs a multiple level service matching framework. The framework takes advantage of similarity algorithm which is proposed by this paper to implement matchmaker, and make use of service ontology to test maching effect of framework.