Quality Prediction of Web Services Based on a Covering Algorithm

Service-oriented architecture (SOA) is widely used, which has fueled the rapid growth of Web services and the deployment of tremendous Web services over the last decades. It becomes challenging but crucial to find the proper Web services because of the increasing amount of Web services. However, it proves unfeasible to inspect all the Web services to check their quality values since it will consume a lot of resources. Thus, developing effective and efficient approaches for predicting the quality values of Web services has become an important research issue. In this paper, we propose UIQPCA, a novel approach for hybrid User and Item-based Quality Prediction with Covering Algorithm. UIQPCA integrates information of both users and Web services on the basis of users’ ideas on the quality of coinvoked Web services. After the integration, users and Web services which are similar to the target user and the target Web service are selected. Then, considering the result of integration, UIQPCA makes predictions on how a target user will appraise a target Web service. Broad experiments on WS-Dream, a web service dataset which is widely used in real world, are conducted to evaluate the reliability of UIQPCA. According to the results of experiment, UIQPCA is far better than former approaches, including item-based, user-based, hybrid, and cluster-based approaches.