EXACT KNOWLEDGE HIDING IN TRANSACTIONAL DATABASES

Abstract

The hiding of sensitive knowledge in the form of frequent itemsets, has gained increasing attention over the past years. This paper highlights the process of border revision, which is essential for the identification of hiding solutions bearing no side-effects, and provides efficient algorithms for the computation of the revised positive and the revised negative borders. By utilizing border revision, we unify the theory behind two exact hiding algorithms that guarantee optimal solutions both in terms of database distortion and side-effects introduced by the hiding process. Following that, we propose a novel extension to one of the hiding algorithms that allows it to identify exact hiding solutions to a much wider range of problems (than its original counterpart). Through experimentation, we compare the exact hiding schemes against two state-of-the-art heuristic algorithms and demonstrate their ability to consistently provide solutions of higher quality to a wide variety of hiding problems.