K-ARQ : K-Anonymous Ranking Queries

Abstract

With the advent of an unprecedented magnitude of data, top-k queries have gained a lot of attention. However, existing work to date has focused on optimizing efficiency without looking closely at privacy preservation. In this paper, we study how existing approaches have failed to support a combination of accuracy and privacy requirements and we propose a new data publishing framework that supports both areas. We show that satisfying both requirements is an essential problem and propose two comprehensive algorithms.We also validated the correctness and efficiency of our approach using experiments.