【摘要】 为了提高分布式网络中各节点信任评价的准确度,提出了一种基于个体经验的信任模型.该模型通过引入经验因子和相对经验因子的方法,建立了新的信任评价体系.这种新的信任评价体系考虑了个体节点的差异问题,在计算节点的信任值时考虑了节点间的交互历史,这在一定程度上解决了由于节点的非对称性而导致的信任评价不准确的问题.算法分析表明:新模型能够针对不同的个体节点,采用不同的最大容忍评价偏差,并且对个体节点的反馈可信度进行更新时,采用不同的更新值,体现了节点的个性化特征,使信任评价更加准确合理.此外,所提出的新算法能够运用到多种信任模型中,具有很好的可扩展性.更多还原

【Abstract】 To improve the accuracy of node trust evaluation in a distributed network,a trust model based on the experience of individuals is proposed,which establishes a new trust assessment system by introducing the experience factor and the comparative experience factor.The new evaluation system considers the differences between individuals and interactive histories between nodes,which solves the problem that nodes have inaccurate assessments due to the asymmetry of nodes to a certain extent.The algorithm analysis indicates that the new model uses different deviating values of tolerance evaluation for different individuals and uses different updating values embodying node individuation when updating feedback credibility of individuals,which evaluates the trust value more reasonably and more accurately.In addition,the proposed algorithm can be used in various trust models and has a good scalability.更多还原