(k, ε)-Anonymity: An anonymity model for thwarting similarity attack

Abstract

Existing anonymity models rarely consider the semantic similarity between sensitive values, so they cannot thwart similarity attack. To solve the problem, this paper proposes a (k, ε)-anonymity model which requires that each equivalence class in anonymous dataset satisfy k-anonymity constraints. At the same time, any two sensitive values in the same equivalence class are not ε-similar. The paper also proposes a (k, ε)-KACA algorithm. Experimental results show that the anonymous data satisfy(k, ε)-anonymity has higher diversity than that satisfy k-anonymity model, so (k, ε)-anonymity model can protect privacy more effective than k-anonymity model.