Qos-based Semantic Web Services Discovery System

With Web services technology's rapid development, the application of Web services is becoming more and more widely. The aim of Web services is to complete the dynamic and loosely coupled application through making themselves that may be provided by different people or organizations distributed in different areas interact and integrate. However, how to find the best service among such a number of Web services in order to meet user needs is becoming an urgent problem because of an increasing number of Web services published on the Internet. So the matching and discovery issues of Web services have become a hotspot.Traditional Web service discovery technology based on keywords is only limited to matching the syntax without description of semantic, which either the accuracy of discovery results or efficiency of user feedback on the results can not meet the current requirements; Web Services Description Language WSDL which is used to describe Web services is only limited to syntax level, which can only provide basic information of invoking services without semantic description of service functions; UDDI which is used to query and discover Web services does not take into account of the problems of how to filter and select the best services that can best meet user needs, that is the QoS requirement.First, in order to solve the problems presented above, this paper studies the semantic Web services discovery mechanism and proposes semantic Web services discovery system model based on QoS.In this model, we use OWL-S to describe semantic Web services and extend OWL-S to increase the description of quality of service. Second, we use OWL-S/UDDI Matchmaker to extend the UDDI registry in order to support semantic description. In the semantic Web services matching algorithm we change the description of semantic Web service's Input and Output matchmaking to the computing of semantic distance in domain ontology base and propose computing formula of concept similarity. Third, the paper describes algorithms of matchmaking. In the QoS restriction matching process, we select Web services of high performance using cost performance, and users also can select Web services that met their needs. The paper can sort the matched results as the desire of requestor in order to meet different demand from requestor. This paper also designs Web service center which is to store the published services'OWL-S files and ontology repository which is to store ontology information.In the end of the paper we make a conclusion of the proposed algorithm and system and present research directions for future work.