Datafly算法是数据发布环境下保护数据隐私的一种k-匿名方法,实现k-匿名时只对准标识符属性集中属性值种类最多的属性进行归纳.当准标识符属性集中只有一个属性的取值多样而其他属性取值具有同质性时,该算法可行.实际应用中数据的取值却往往不具有这种特点.针对这个问题,提出一种自底向上的支持多属性归纳k-匿名算法,并对该算法进行实验测试,结果表明该算法能有效降低原始数据的信息损失并能提高匿名化处理效率.%Datafly algorithm is an k-anonymity method for protecting data privacy in privacy preserving data publishing, the most frequent attribute of quasi-identifier attributes is generalized when realizing k-anonymity. Datafly algorithm can be executed when the values of an attribute of quasi-identifiers are diversity and the values of the other attributes are homogeneity. However, the character is impossible in practical applications. According to the problem, an bottom-up generalization algorithm for supporting multi-attribute is building. Experimental results demonstrate that the developed algorithm is efficient for solving information loss and elapsed time.
展开▼