An efficient method to implement data private protection for dynamic numerical sensitive attributes
Xiaoling Xia, Qiang Xiao, Wei Ji · 2012
Recently, great attention has focused on the data private protection for dynamic publishing data in the database community. A novel concept named m-invariance has solved the difficulty of updating data in the consecutive generalized tables, but the little concern about the performance and numerical sensitive attributes has restricted the utilities of this cunning method. In this paper, by analyzing the progress of the m-invariance, we reveal its defect, introduce a definition “selective factor” to decrease randomness of m-invariance, and propose a method named NCm-invariance (NSA to CSA m-invariance) to convert numerical sensitive attributes (NSA) to categorical sensitive attributes (CSA), to tackle the problem.