Predicting the Quality of Web Services Based on User Stability

Abstract

The development of web services and web APIs offers software developers great opportunities for choosing reliable services. However, the quality of these web services are often not available. Existing quality of web service prediction methods adopts the recommender system related techniques to predict the service quality. In these approaches, the behaviour of service invokers do not change. In reality, the service invokers network conditions are changing all the time. This fact inspires us jointly to consider the stability of service invokers network environment when building a prediction model. In specific, a reliability model is adopted for stability calculation and a recommendation algorithm is proposed in this paper. The advantages of our proposed algorithm is confirmed via experiments on a real-life quality of web service data set and comparison with existing quality of web service predicting algorithms.