A novel approach for QoS prediction based on Bayesian combinational model

As an important factor in evaluating service, QoS (Quality of Service) has drawn more and more concerns with the rapid increasing of Web services. However, due to the great volatility of services in Mobile Internet environments, such as internet of vehicles, Web services often do not work as announced and thus cause unacceptable problems. QoS prediction can avoid failure before it takes place, which is considered a more effective way to assure quality. However, Current QoS prediction approaches neither consider the highly dynamic of Web services, nor maintain good prediction performance all the time. Consequently we propose a novel Bayesian combinational model to predict QoS by continuously adjusting credit values of the basic models so as to keep good prediction accuracy. QoS attributes such as response time, throughput and reliability are used to validate the proposed model. Experimental results show that the model can provide stable prediction results in Mobile Internet environments.