A Novel Reputation Model For Web Services Selection With Raters' Sensitivity

We propose a service's reputation model in service-oriented environments. The model is built on the base of raters' behavior showing on current rated service and other similar services. Both a rater's credibility and sensitivity are considered in our model at the same time. Furthermore, the latter is first put forward distinguishing among the services to give constructive suggestions for new users. And temporal characteristics of the ratings are considered throughout the model establishment. Experiments are benchmarked to demonstrate that all the factors play an important role in the final reputation of a service. Specially, it is also observed that sensitivity is excellent to increase the distinction of reputations among different services.