面向个体和敏感属性值的匿名数据发布  

ANONYMOUS DATA PUBLISHING FROM THE PERSPECTIVE OF INDIVIDUAL AND SENSITIVE ATTRIBUTE VALUES

作　　者：骆永成[1,2] 乐嘉锦[2] 陈德华[2] 

机构地区：[1]东华大学图书馆,上海201620 [2]东华大学计算机科学与技术学院,上海201620

出　　处：《计算机应用与软件》2012年第6期5-7,75,共4页Computer Applications and Software

基　　金：国家自然科学基金项目(61070031)

摘　　要：针对目前数据发布方法不能有效处理不同个体隐私保护需求的问题,依据个体隐私自治的原则,从面向个体和敏感属性值角度,提出一个敏感数据发布的个性化匿名发布模型和基于泛化技术的启发式算法。通过Adult数据实验,验证了算法的可行性。与Basic Incognito和Mondrian相比,信息损失少,算法性能良好。While current data publishing methods can not effectively deal with different issues of individual privacy protection needs,this paper,based on the autonomy principle of individual privacy,puts forward a personalised anonymous publishing model applied to release the sensitive data and a generalisation-based heuristic algorithm,both are from the perspective of individual and sensitive attribute values.Finally,the Adult data experiments verify the feasibility of this algorithm.These show that it has less loss of information than the algorithm of Basic Incognito and Mondrian,and has good execution performance.

关 键 词：数据发布 个性化 泛化 匿名模型 

分 类 号：TP309[自动化与计算机技术—计算机系统结构]