A framework for reputation model based on time and trust aware Heteroscedastic Gaussian process

The evaluation of reputation of Web services on the user side is of paramount importance while selecting a web service. Due to dynamic nature of environment where web services are deployed and accessed, evaluation of service trust based on reputation feedback is highly time dependent. The correct evaluation is always influenced by the trustworthiness of the source of feedback as perceived by the evaluator, and also by the correctness (precision) of the feedback. The paper presents, a framework for a time and trust aware mechanism for inference and prediction of reputation using Heteroscedastic Gaussian process regression (HGPR) model. The model accepts, as input, the reputation feedback from third party sources and predicts the real reputation value.