INTERNATIONAL JOURNA L OF ENGINEERING SCI ENCES & RESEARCH TECHNOLOGY Privacy Preserving Data Publish ing: Using Overlapping Slicing a nd Attribute Partitioning
R. Sugumar, Vel Tech · 2013
Privacy preserving publishing is the kind of techni ques to apply privacy to collected vast amount of The data publication processes are today still very difficult. Data often contains personally identifi able information and therefore releasing such data may result in pri vacy breaches; this is the case for the examples of microdata, e.g., and medical data. The proposed techniques in this p roject accelerate accessing speed of user as well a s applying privacy to collected data. Several anonymi zation techniques were designed for privacy preserv ing data publishing. Recent work in data publishi ng has shown that generalization losses considerabl e amount of information, especially for high dimensional data. Bucketization, on the other hand, does not prevent membership disclosure. I propose an overlapping slicing method for handling high -dimensiona l data. By partitioning attributes into more than one column, we protect privacy by br eaking the association of uncorrelated attributes a nd preserve data utility by preserving the association between highly correlated attributes. This technique releas es mo correlations thereby, overlapping slicing preserves better data utility than generalization and is mor e effective than bucketization in workloads involving the sensitive attribute.