Research On Dynamic Trust Management Mechanism Based On User Behavior Under Cloud Computing Environment

With the fast development of cloud computing, more and more user starts to usethe cloud service. And with the explosive growth of user data, it has been a veryimportant research problem to make sure that cloud user can use resource of cloudplatform securely and efficiently in the cloud computing security research.Focusing on this problem, this paper presents a dynamic trust managementmechanism based on user behavior from the trust management perspective. Thismechanism consists of trust management model and the prototype system of thismodel. Firstly, we collect the data of cloud user behavior by the terminal template ofcloud platform. Secondly, analyzing each data property by the data mining algorithmK-Prototype and KNN in order to get different data clusters, and calculate the weightof each data prototype by entropy-weight algorithm. Thirdly, we propose a trustalgorithm which combines user history trust degree with analyzing result of datamining to calculate user trust value and evaluate user trust degree. In the research, wealso finished the prototype system of this trust model. This system can realize thebasis function of the trust model and calculate user trust value by collecting userbehavior data.This dynamic trust management mechanism can evaluate user final trust degreeby user behavior data, user history trust degree and each module in the trust model.From the compare of the experiments in this paper, we can get that this model canreflect the changes in the realistic into the computing result, thus to realize dynamicmanagement of user trust. This paper provides a reference to the research of cloudtrust management, and the research of this mechanism has a certain theoreticalsignificance and application value.