Service-evaluation-based resource selection for cloud manufacturing

With the development of cloud computing, increasing attention has shifted from traditional manufacturing toward cloud manufacturing. The distinguishing feature of cloud manufacturing is that the resource designated for a given manufacturing service is always massive, complex, and heterogeneous. Cloud manufacturing also involves a high degree of user participation and user diversity. This article presents a method of resource selection based on service evaluation, which combines predictive evaluations and recommended evaluations. Predictive evaluations are based on the user's historical evaluations of a given service, and these evaluations may be weighted differently depending on the time when the service occurred. Recommended evaluations are given by recommenders who are selected by a two-step process. These evaluations are weighted according to their similarities and objectivities. The results of numerical experiments show that the proposed method performs better than previous methods.