PDMP:εk个性化数据脱敏保护方法  

PDMP:εk personalized data masking protection

作　　者：刘振鹏[1,2] 孙静薇 王烁 王文胜 尹文召 张彬 Liu Zhenpeng;Sun Jingwei;Wang Shuo;Wang Wensheng;Yin Wenzhao;Zhang Bin(School of Electronic Information Engineering,Hebei University,Baoding Hebei 071002,China;Center of Information Technology,Hebei University,Baoding Hebei 071002,China;School of Cyber Security&Computer,Hebei University,Baoding Hebei 071002,China)

机构地区：[1]河北大学电子信息工程学院,河北保定071002 [2]河北大学信息技术中心,河北保定071002 [3]河北大学网络空间安全与计算机学院,河北保定071002

出　　处：《计算机应用研究》2020年第10期3068-3070,3082,共4页Application Research of Computers

基　　金：河北省自然科学基金资助项目(F2019201427);国家教育部“云数融合科教创新”基金资助项目(2017A20004)。

摘　　要：通过对k-匿名数据脱敏和l-多样性匿名数据脱敏模型分析,提出一种更加有针对性的个性匿名保护方法(PDMP)来解决数据泄露问题。PDMP根据不同的敏感程度来找到不同的k值,通过相应的约束方法对数据进行泛化从而实现敏感数据的脱敏,减少真实数据的攻击率,更好地实现隐私数据的保护。实验表明,该方法降低了信息敏感度的同时,满足了个性化需求,有效地提高了数据隐私的安全性。Through the desensitization of k-anonymous data and the desensitization model of l-diversity anonymous data,this paper proposed a more targeted personality data masking protection method(PDMP)to solve the data leakage problem.PDMP found different k values according to different sensitivity levels,and generalized the data through corresponding constraint methods to achieve desensitization of sensitive data and reduced the attack rate of real data to protect private data better.Experiments show that the proposed method reduces the information sensitivity and it satisfies the individualized requirements and effectively improves the security of data privacy.

关 键 词：个性化 K-匿名 泛化 l-多样性 隐私保护 

分 类 号：TP309.2[自动化与计算机技术—计算机系统结构]