开放式网络信任计算中多影响因子的分析  

Analysis of multi-factor in trust evaluation of open network

作　　者：赵斌[1,2] 何泾沙[1,3] 张伊璇[1] 翟鹏[2] 刘公政 周世义 

机构地区：[1]北京工业大学软件学院,北京100124 [2]济宁学院计算机科学系,山东曲阜273155 [3]北京经开投资开发股份有限公司信息技术部,北京100176

出　　处：《山东大学学报（理学版）》2014年第9期103-108,共6页Journal of Shandong University(Natural Science)

基　　金：国家自然科学基金资助项目(61272500);北京市自然科学基金资助项目(4142008);北京市委;市政府重点工作及区县政府应急项目(z131100005613030);山东省自然科学基金资助项目(ZR2013FQ024);山东省科学技术发展计划项目(2012RKA08002;2013RKA08007);山东省高校科技计划项目(J12LN70;J14LN80)

摘　　要：信任模型是解决开放式网络环境中信任问题的有效方式,信任度量是可信管理中亟待解决的关键问题。根据信任模型设计原则,针对信任计算中动态适应能力和信任有效聚合不足等问题,在提出的信任计算模型中引入奖惩因子、推荐实体的评价可信度和平衡权重等多影响因子来提高信任的可靠程度和精度,并给出了各影响因子的设计。最后,通过Matlab软件仿真实现了奖惩函数和平衡权重的选取分析。仿真实验结果表明,该信任计算模型有较好的聚合速度。Trust model is an effective way of resolving trust problems in open networks in which trust evaluation is a key issue to be addressed in trust management.According to the design rules of trust models,the problems of lacking dynamic adaptability in trust evaluation and the lack of effective methods for the aggregation of trust,the multiple fac-tors that include the bonus-penalty factor,evaluation reliability of the recommending entities and balanced weight were introduced into trust evaluation to improve the realiblity and accuracy of trust.The design of the bonus-penalty function was also presented.Finally,the balanced weight factor and an analysis was carried out through the Matlab software.The experimental results show that the proposed trust model can achieve a better aggregation speed.

关 键 词：开放式网络 信任计算 影响因子 仿真分析 

分 类 号：TP393[自动化与计算机技术—计算机应用技术]