A Referral-Based QoS Prediction Approach for Service- Based Systems

Abstract

During development of service-based systems (SBS), the quality of services (QoS) plays an important role at helping select more suitable services. There have been several QoS prediction approaches proposed, however, their prediction accuracy is low when there are few historical records in the application environment. In this paper, we propose a new QoS prediction method based on a virtual platform methodology. The method first constructs a virtual platform based on Gaussian distribution regarding stability and performance of services. With the platform, a referral-based QoS prediction method has been developed to improve prediction accuracy. The experimental results indicate that our method outperforms previous approaches, achieving higher prediction accuracy, especially when there are few historical records available.

A Probability Distribution and Location-aware ResNet Approach for QoS Prediction

“…In different network environments, QoS values generated by different users may vary greatly even if the invoked service is the same one. Therefore, it is meaningful to predict QoS values of candidate services before the user invokes a service, which can help the target user distinguish the most suitable service among different functionally equivalent services according to the predicted QoS results [1,2,3,4,5,6]. At present, QoS value has become a pivotal criterion for service selection and service recommendation, and QoS prediction has also been applied in plenty of service recommendation systems.…”