基于聚类的(L,K)-匿名

K-匿名是解决数据隐私的关键技术,成为近年来研究热点.目前对K-匿名的研究大多依赖预定义的泛化层次,泛化后的数据有很大的数据损失,并且没有考虑到匿名后的可信属性缺乏多样性导致的隐私信息泄漏.本文针对K-匿名存在的上述问题,提出了一种在K-匿名之上的(L,K)-匿名模型,将聚类的方法应用(L,K)-匿名模型上,并给出了基于聚类分析的(L,K)-匿名算法,实验显示该方法能有效的消除K-匿名后可信属性的信息泄漏,增强数据发布的安全性.

(L,K)-anonymity based on clustering

K-anonymization techniques are a key component of any solution to data privacy and have been the focus of intense research in the recent years. Current solutions, however, suffer from one or more of the following limitations: reliance on pre-defined generalization hierarchies; generation of anonymized data with high information loss and the inference channel arising from lack of diversity in the sensitive information. In this paper a new privacy protection method beyond K-anonymity called (L, K)-anonym-ity and uses the idea of clustering is introduced. The algorithm of (L,K)-anonymity based on clustering is provided, the experiments shows that algorithms can eliminate the information disclosure after K-anonymity efficiently and can enforce the data security.