(BWTD)_(M):一种有效的信誉攻击防御策略  

(BWTD)_(M):An effective reputation attack-resistance strategy

作　　者：周尤明 Zhou Youming(College of Logistic and Information Engineering,Huzhou Radio and Television University,Huzhou 313000,China)

机构地区：[1]湖州广播电视大学物流与信息工程学院,浙江湖州313000

出　　处：《南京理工大学学报》2023年第1期81-89,共9页Journal of Nanjing University of Science and Technology

基　　金：国家自然科学基金(92046002);浙江省自然科学基金(LY20F030002)。

摘　　要：为解决目前信誉系统的防御能力与信誉量化准确度的矛盾问题,提出了(BWTD)_(M)信誉攻击防御策略。第一,在审查团构建上,为了针对性提高防御能力,按照用户不同商品需求类型配置不同的黑白名单,即多维(Multiple,M)黑白(Black,B;White,W)名单;第二,为解决因多维黑白名单导致数据稀疏从而影响信誉量化准确度的问题,在信任(Trust,T)量化上,评价者信任从社交、用户行为和能力等多维数据进行量化,同时在不信任(Distrust,D)量化上,评价者不信任通过多维恶意评价数据进行针对性量化;第三,在信誉量化上同时考虑第一和第二点。试验表明该策略在信誉评价的准确度和抗攻击能力上明显优于现有防御策略。研究结果有利于对信誉攻击进行防御。To cope with the contradiction between the attack-resistance capacity and the accuracy of the reputation quantification,a reputation attack-resistance strategy called(BWTD)Mis proposed(Multiple,M;Black,B;White,W;Trust,T;Distrust,D).Firstly,in terms of censor formations,to pertinently enhance the attack-resistance capacity,users are equipped with multifarious different black-white lists according to their different products requests,that is,multi-dimensional black-white lists.Secondly,to cope with the data sparseness problem(affects the the accuracy of the reputation quantification)caused by multi-dimensional black-white lists,in terms of the trust quantification,evaluator trust is quantified in multiple dimensions including social data,user behavior data and capability data.Meanwhile,in terms of the distrust quantification,evaluator distrust is pertinently quantified through multi-dimensional malicious evaluation data.Thirdly,in the process of reputation quantification,the first point and the second point are concerned simultaneously.The experiments show that the presented strategy is better than the current resistance strategies in both reputation evaluation precision and reputation attack-resistance capabilities.The research results are helpful to the reputation attack-resistance.

关 键 词：攻击防御 信任与不信任 恶意评价特征 信誉评价 黑白名单 

分 类 号：TP309[自动化与计算机技术—计算机系统结构]