Privacy-preserving data analysis with differential privacy

Abstract

There is a growing demand for statistical data analysis as it is being adopted by many businesses for a variety of purposes. For that reason, there has been a lot of attention recently focused on the concept of differential privacy as a standard model for ensuring privacy when personal information is used and sold for secondary purposes. In this work, we explore the available privacy-preserving mechanisms and test one of the more mature techniques in an experiment. In the end, we present the findings and conclude on the practicality and accuracy of the technique.