Personalized privacy protection for transactional data

Abstract

Privacy protection in publication of transactional data is an important problem. However,the bulk of existing methods focus on a universal approach that exerts the same amount of preservation for all users and all sensitive values, without considering different requirements of users and different sensitivity of concrete attribute values. Motivated by this, we introduce a framework which provides personalized privacy protection based on the form of bipartite graphs via partition and generalization. Our approach can preserve privacy of sensitive associations between entities and retain the largest amount of nonsensitive associations to provide better data utility. Experiments have been performed on real-life data sets to measure the accuracy of answering aggregate queries. Experimental results show that our approach offer strong tradeoffs between privacy and utility.