多敏感属性K-匿名模型的实现  

Implementation of K-anonymous Model with Multi-sensitive Attributes

作　　者：王静[1] 闫仁武[1] 刘亚梅[1] 

机构地区：[1]江苏科技大学计算机科学与工程学院,镇江212003

出　　处：《计算机与数字工程》2017年第7期1368-1372,共5页Computer & Digital Engineering

摘　　要：L-MDAV算法结合L-多样性模型和MDAV微聚集算法来实现K-匿名模型,但是L-MDAV算法只考虑了单一敏感属性下的约束,而在实际发布的数据中不可能只有一个敏感属性,论文在该算法的基础上提出了基于多敏感属性的MDAV算法,文中仅以两种敏感属性为例,因此算法命名为(L1,L2)-MDAV算法。论文在根据用户不同敏感属性的不同保护需求下为用户个性化定义敏感属性的敏感度。实验结果表明,论文提出的算法相比于L-MDAV算法能够更好地保护隐私数据安全。L-MDAV algorithm combined with L -d ive rs ity model and MDAV algorithm to implement the K-anonymous mod-el, but L-MDAV algorithm only considers the single sensitive attribute, in the actual , released data may not have only one sensitive attribute, so, on the basis of the algorithm the article propose a MDAV algorithm with multi-sensitive attributes. In this paper, with only two sensitive attributes as an example, so the algorithm named（ L1,2 ）-MDAV algorithm. According to user’s protection for d if-ferent sensitive attributes have different requirements to define the different values of L . As the experimental results show, the pro-posed algorithm can protect the privacy of privacy data perfectly compared with L -MDAV algorithm.

关 键 词：多敏感属性 K-匿名 L-多样性 微聚集算法 

分 类 号：TP3-05[自动化与计算机技术—计算机科学与技术]