社交网络中攻击背景下个人隐私泄露度量研究

Leakage metrics of personal privacy under attack background in social network

社交网络中隐私泄露受多种因素的影响,针对不能有效度量隐私泄露的问题,尤其是在攻击者拥有背景知识的时候,必须综合考虑背景知识来设计隐私泄露度量方法.构建一个个人隐私泄露多维分析模型,对隐私泄露的影响因素进行较全面的分析,结合泄露指标的计算,推导出影响较大的因素.提出一种背景知识分类和量化方法,能划分关键背景知识和非关键背景知识;在此基础上提出一个攻击背景下隐私泄露度量信息熵模型和隐私泄露的综合度量方法,解决隐私泄露有效度量的问题.所提出的模型和背景知识量化方法可为攻击背景下隐私泄露风险分析与评估提供可行的计算基础,实验验证和分析结果表明度量模型是可行的、有效的.

Aiming at the problem of accurate measurement of privacy leakage in social networks,especially when attackers has attack background knowledge,while privacy leakage is affected by many factors,a privacy leakage metrics method considering background knowledge should be considered comprehensively.A multi⁃dimensional analysis model for personal privacy leakage is proposed in this paper.Influencing factors of privacy leakage can be analyzed comprehensively.The greater influencing factor is deduced through calculation of leakage index.A method of classification and quantification of background knowledge is proposed,which can distinguish key background knowledge and non key background knowledge.On this basis,an information entropy model and a comprehensive measurement method of privacy leakage are proposed to solve the problem of measuring privacy leakage accurately.The proposed model and background knowledge quantification method can provide a favorable support for the quantitative research on the impact of attack background knowledge.Experimental verification and analysis results show that the measurement model is feasible and effective.