A Trust Model for Secure and Reliable Cloud Service Systems

With the ever-increasing popularity of cloud computing technology, many enterprises and organizations have migrated their business applications to the cloud platform, thus forming a cloud service system that hosts various cloud services. A user can browse, select and invoke his/her interested services from the cloud service system and rate the service quality through online feedback or ratings. Such user ratings contribute much to the healthy running of the cloud service system. However, the fake ratings from potential malicious users can affect the fairness of cloud service system's trust model negatively and severely, even misleading the future users' service selection decisions. Therefore, it is becoming a necessity to build a find-grained trust model that can effectively punish the malicious ratings, for the security and reliability of the cloud service system. In view of this challenge, in this paper, we put forward a novel punishment approach for malicious ratings based on three criteria, i.e., cost, profit and risk. More concretely, we analyze each criterion as well as its influencing factors and put forward a unified model that integrates the three criteria quantitatively.