视图的k-匿名化方法

k-匿名是防止链接攻击所造成的发布视图信息泄露的有效方法。在实际应用中往往会同时发布多个视图，如何使视图组满足k-匿名约束亟待解决。首先，分析了视图分别进行概括、保持数据一致性、最小概括情况下，如何使视图组满足k-匿名约束；然后，根据上述各种情况，分别给出了独立概括法、联合概括法和属性概括法等视图组k-匿名化算法。实验证明，所提出的算法都可以有效地使视图组达到k-匿名效果，从而保证发布视图的安全。

K-anonymization method for views

K-anonymity is a primary method for guaranteeing the security of views,it can provide privacy protection and prevent information disclosure induced by joining attack.In practical application,it is usual that the data owner will publish multiple views at the same time,so,how to make the whole view set satisfy k-anonymity constraint become the most imperative problems in the research of k-anonymity.Previous researches about k-anonymity have concentrated on k-anonymization algorithm for single view,but these algorithms can not be adopted by views directly.At first,the method how to make views achieving k-anonymity constraint is analyzed under several cases,such as generalizing view separately,preserving dada consistency,minimal generalization.Then,based on the analysis under the several cases,the k-anonymization algorithms for views,such as independent generalization,association generalization,attribute generalization,are proposed respectively.The experiment results show that each k-anonymization algorithm for view set can make view set satisfy k-anonymity constraint effectively.So,they can guarantee the security of the views effectively in the process of publishing views.