Location-Based Two-Phase Clustering for Web Service QoS Prediction

Quality-of-Service (QoS) is widely employed for describing non-functional characteristics of Web services. QoS prediction for Web services is a hot research problem in service computing. Collaborative Filtering (CF) has been widely used as a prediction technique. However, CF may suffer from data scarcity problem, which causes unreliable prediction. In this paper, we propose a novel location-based two-phase clustering QoS prediction method (LCQP). The two-phase clustering process respectively clusters users and services into the corresponding user clusters and service clusters according to their location information. Then we can apply the CF method by the clustering information instead of individual user and service information. So that there is more information to be used and it will contribute to promote the prediction accuracy. The experimental results showed that our approach could achieve not only higher prediction accuracy, but also lower computation time than other CF-based methods.