Extended K-anonymity models against attribute disclosure

Abstract

P-sensitive K-anonymity model has been recently defined as a sophistication of K-anonymity. This new property requires that there be at least P distinct values for each sensitive attribute within the records sharing a combination of key attributes. However, as shown in this paper, it may not protect sensitive information in some way. In this paper, we empirically investigate two enhanced K-anonymity models. Instead of publishing original specific sensitive attributes, the new models publish the categories that the sensitive values belong to. We propose a top-down approach to implement two enhanced models and show in the comprehensive experimental evaluations that the two new introduced models are practical in terms of effectiveness and efficiency.