A Semantic Web Service Retrieval Approach that Combines Semantic Matching With Quality of Service Matching

Abstract

The study objective is to propose a new web service search system that exploits the information structure of OWLS documents which consists of information from service history class, service model class, and basic service class. This information is used to construct indexes and their individual weights, which are used to compare the level of similarity between a semantic query from a user and each web service by a vector space model. Then, the result is used in a web service selection procedure based on formal concept analysis. A web service is selected through the structure of a concept lattice of Qos consisting of service availability and response time. Experimental results show that this proposed system provided an average accuracy of 71.9%. Moreover, it can provide alternative web services that are closely related to the query for the user to have flexibility in utilizing the search results.