基于最大叶子子树优先策略的多敏感属性保护方法  

A method for protecting multi-sensitive attributes based on maximal-leaf-subtree-first strategy

作　　者：祁瑞丽[1] 王可[2] 郭学涛[1] 李金才[1] 唐军军[1] 刘国华[1] 

机构地区：[1]燕山大学信息科学与工程学院,河北秦皇岛066004 [2]大连理工大学电子与信息工程学院,辽宁大连116000

出　　处：《燕山大学学报》2009年第5期433-437,443,共6页Journal of Yanshan University

基　　金：国家自然科学基金资助项目(60773100);国家"十一五"科技支撑计划资助项目(2006BAK05BO2);河北省自然科学基金资助项目(F2009000475)

摘　　要：首先将多敏感属性隐私保护问题转化为多敏感属性L-多样性问题,然后给出了多敏感属性树构造方法及最大叶子子树优先策略,在此基础上提出了一个多敏感属性保护算法。最后,通过实验对算法进行了验证和分析。结果表明,该方法能有效地保护数据隐私,减少信息泄露。First, the multi-sensitive attributes privacy protecting problem was transformed into the multi-sensitive attributes L-diversity problem. Then, a method of construction for the multi-sensitive attributes tree and the maximal-leaf-subtree-first strategy were presented. Based on this, a multi-sensitive attributes protecting algorithm was proposed. Finally, the algorithm was confirmed and analyzed through experiment. The results show that the method can effectively protect data privacy and reducing information disclosure.

关 键 词：隐私保护 多敏感属性树 最大叶子子树 L-多样性 

分 类 号：TP309.2[自动化与计算机技术—计算机系统结构]