A Computational Trust Model Based on Indirect Information

Trust is a focus of security issues in the diverse, large scale, open environment. Currently, there are many trust models, which require historical information about past behavior of specic agents to determine the trust relationships, but the information about the specic agents is not always available. To solve this problem, the local information, which users need to evaluate the trustworthiness of a strange service provider (target agent), is taken into account. In this paper, we put forward a computational trust model. Firstly, the agents whose information is similar to the target agent are obtained from the users’ local information through the collaborative ltering algorithm. Secondly, we propose a method about the synthesis of trust with timeliness of trust. Finally, simulation results prove that the proposed method based on users’ local information through collaborative ltering algorithm can signicantly improve the accuracy of the assessment.