证据信任模型中的信任传递与聚合研究  

Research on trust transitivity and aggregation in evidential trust model

作　　者：蒋黎明[1] 张琨[1] 徐建[1,2] 廖俊[1] 张宏[1] 

机构地区：[1]南京理工大学计算机科学与技术学院,江苏南京210094 [2]南京大学计算机软件新技术国家重点实验室,江苏南京210094

出　　处：《通信学报》2011年第8期91-100,共10页Journal on Communications

基　　金：国家自然科学基金资助项目(60903027;61003210);江苏省自然科学基金资助项目(SBK201022379);南京理工大学自主科研专项计划基金资助项目(2010ZYTS035;2010GJPY056);高等学校博士学科点专项科研基金资助项目(20093219120024)~~

摘　　要：现有信任模型在信任传递和聚合方面存在一定的不足:首先是信任传递过程中缺少对直接推荐实体的反馈信任度的有效度量;其次是信任聚合过程中因缺少对推荐链之间依赖关系的有效处理而存在推荐信息的损耗或重复计算等问题。为了解决这些问题,提出了一种新的证据信任模型,结合D-S证据理论和图论方法有效解决了信任传递的可靠性和信任聚合的准确性等问题。仿真实验表明,与已有模型相比,所提模型具有更强的抑制各种策略欺骗及共谋行为的能力,在信任度量准确性方面也有较大提高。Current trust models had some certain disadvantages in dealing with trust transitivity and trust aggregation.Firstly,these models ignored the feedback trust measurement on the direct recommender in the process of trust transitivity.Secondly,the models contained the problems of information loss or repetitive calculation for the limits in analysis and dis-posal of the dependent relationships among referral chains.A new evidential trust model was proposed,and the problems existed in current methods for transferring and aggregating trust relationships were addressed by combing D-S evidence with graph theory.It can be seen from the simulation results that compared with existing trust models,the proposed model is more robust on defending malicious attacks for various strategy cheating and collusion,and has more remarkable en-hancements in the accuracy of trust measurement.

关 键 词：D-S证据理论 信任传递 信任聚合 信任子图 

分 类 号：TP393[自动化与计算机技术—计算机应用技术]