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