A graph-based QoS prediction approach for web service recommendation

With the development of the Internet, the recommendation based on Quality of Service(QoS) is proven to be an efficient way to deal with the ever-increasing web services in both industry and academia. However, it is hard to make an accurate recommendation using sparse QoS data, which makes QoS prediction a growing concern in the context of web service recommendation. In this research, a novel Graph-based Matrix Factorization approach(GMF) is proposed for QoS prediction. First, a concept of integrated-graph is put forward to consolidate multi-source information from user-aware context and service-aware context, and to deep mine potential relationships based on QoS matrix. Furthermore, the integrated-graph is divided into several sub-graphs by cutting insignificant edges to reduce noises and strengthen interactions between users and services. Based on the local information of each sub-graph and the global information of integrated-graph, a Gaussian Mixture Model(GMM) of QoS value is built as a fusion method to combine local and global information adaptively and to complete final QoS prediction. The extensive experimental analysis on a publicly available dataset indicate that our graph-based method is both accurate and practical.

Automated summary

With the development of the Internet, the recommendation based on Quality of Service (QoS) has become increasingly important in dealing with web services. The research proposes a Graph-based Matrix Factorization approach (GMF) for QoS prediction, which consolidates multi-source information and uses a Gaussian Mixture Model (GMM) to combine local and global information for accurate predictions. Extensive experimental analysis on a publicly available dataset demonstrates the accuracy and practicality of the proposed method.