Publication

Abstract : Web service are basically software components that support interoperable interaction between machines over a network. With booming number of WS, it's difficult for users to identify the best quality services. Existing researchers focus mainly on selection of WS based on the functional requirements. This paper proposes a system to select an optimal WS by predicting its QoS value based on dynamic non-functional quality factors response time, throughput and the static factor security of the WS. Prediction result can be used in recommendation systems to select services with optimal QoS performance among a large volume of service candidates. Users and web services are clustered to make prediction of response time and throughput of WS. Security specifications of the web service and their vulnerabilities are used for prediction of security factor of the WS. Finally, QoS of the WS is predicted by aggregation of the predicted quality values for security, response time and throughput. The QoS prediction of the system reveals optimal results as it takes both dynamic and static quality factors of WS. © International Science Press.