The Estimation of Trustworthy of Grid Services Based on Neural Network

Though research on the Grid Services has progressed at a steady pace, its promise has yet to be realized. One major difficulty is that, by its very nature, the Grid Service is a large, uncensored system to which anyone may contribute. This raises the question of how much credence to give each source. The concept and definition of trustworthy of Grid Service is given. Estimating trustworthy of Grid Services with the method of Neural Network from the aspect of trustworthy history sequence is proposed. The principle of the method, applicable Neural Network structure, Neural Network constructing, input standardization, training sample constructing, and the procedure of estimating trustworthy of Grid Services with trained Neural Network are presented. Experiments confirm that the methods with Neural Network are feasible and effective to estimate trustworthy of Grid Service, and do not put unreasonable expectations on users. We hope that these methods will help to move the Grid Service closer to fulfilling its promise.