基于多维信任云的信任模型研究

针对主观信任的模糊性和不确定性以及现有的基于云模型的信任模型中粒度粗糙的问题，提出了一种基于多维信任云的信任模型。首先，依据实体间的直接交互经验和交互时间，利用加权逆向云生成算法计算被评估实体的直接信任云；然后，通过评估推荐实体的推荐可信度，计算被评估实体的推荐信任云；最后，综合直接信任云和推荐信任云产生综合信任云，并依此进行可信实体的选择。仿真结果表明，基于多维信任云的信任模型能够有效地识别系统中的各类服务实体，可提高实体间的交易成功率。

Research of trust model based on multidimensional trust cloud

According to the characteristics of fuzziness and uncertainty in subjective trust, and the deficiency of coarse granularity in the existing trust model based on the cloud model, a new trust model based on multidimensional trust cloud was proposed. Based on the feedback and time for direct transactions between the entities, direct trust cloud for every entity could be generated using the weighted backward cloud algorithm. Recommend trust cloud was obtained by integrating recommend information and the recommend credibility of entity was the weight of the recommend information. Finally, the direct trust cloud and recommend trust cloud was synthesized together and the result could be a reference for the object selection. The experimental results show that the multidimensional trust cloud model can detect every kind of service provider effectively and improve the success rate between the entities.