A Novel Dynamic Cloud Service Trust Evaluation Model in Cloud Computing

Abstract

There are many different types of cloud services. How to choose the best cloud service for a specific application is a challenging issue. Using cloud services, users' private data could be exposed to a third party, which might not be trusted. In this paper, a dynamic cloud service trust evaluation model is proposed based on service-level agreement (SLA) and privacy-awareness. First, to make the final trust evaluation value more practical, the proposed model performs a comprehensive trust evaluation, which consists of direct, indirect, and reputation trust. Second, cloud services are divided into five levels based on their service capabilities. By analyzing SLA to determine the quality of service, the user can choose a suitable SLA. In order to protect data security of the cloud service user, we propose a data protection method based on a normal cloud model. Finally, we propose a dynamic trust update mechanism to update the direct trust. Experimental results based on a public dataset show that the proposed model effectively identifies the user community based on service preferences, improves the service requester's satisfaction, avoids malicious interference, and is accurate and feasible.