An Improved Algorithm for K-anonymity

Abstract

K-anonymity is an important model that prevents joining attacks in privacy protecting. Many works have been conducted to achieve k-anonymity. OLA (Optimal Lattice Anonymization) is an efficient full-domain optimal algorithm among these works. The OLA algorithm uses a binary search to traverse the lattice and marks all the k-anonymous nodes, the process of which is time-consuming. In this paper, an improved algorithm based on OLA algorithm is proposed. We firstly computed the product of degree for each node in the lattice and selected the biggest one to judge whether it was k-anonymous; this changed the traversal sequence of binary search. Then, we introduced the conception of support from data mining, and augmented the structure of generalization hierarchy associated with the information of support. As a result there is no need to scan the entire data table repeatedly and we can find all k-anonymous nodes more efficiently. © 2012 Springer-Verlag Berlin Heidelberg.