An Improved Direct Trust Evaluation Algorithm for the Context-aware Trust Model

The traditional trust models that concern single service with constant contexts are inapplicable to deal with the data sparsity problem caused by the diversity of services and contexts. The existing context-aware trust models try to solve this problem by referring to the similarity of services in trust evaluation with low calculation efficiency. In this paper, we propose a Leader-Follower clustering based direct trust evaluation algorithm for the context-aware trust model. The direct trust evaluation for the forthcoming interaction is on the basis of the clusters of trust reference set other than every recent interaction record. The simulation experiment shows that the proposed method can obtain accurate direct trust evaluation result even lacking recent interaction experience for the current service, but the calculation efficiency is higher than the existing context-aware trust evaluation methods.