A Web Service Recommendation Approach Based on Collaborative Filtering

Abstract

With the widespread of Web services, the number of Web services with identical functions is growing sharply. Therefore, it is necessary to recommend Web services to users. This paper proposes a Web service recommendation approach based on collaborative filtering. We first calculate user similarity and service similarity using the historical QoS (Quality of Service) invoking information of Web services. Then, we predict QoS values of services for active users using information provided by similar users and services. Finally, based on the prediction results, we make recommendation for active users. To study the QoS value prediction accuracy of our approach, we conduct real-world experiments. The experimental results show that our approach outperforms other approaches very well.