Differential Privacy Framework: Impact of Quasi-identifiers on Anonymization

Abstract

Due to the high volume of data available with social networking sites and companies, the privacy of the individuals is at a continuous risk. With the help of auxiliary data, the users can be tracked back. It becomes even more necessary to analyse the huge piles of data for research intentions. Hence, protection of privacy is a big concern. To deal with the privacy concerns, numerous privacy paradigms have been proposed to achieve an equilibrium between data utility and privacy. Anonymization of data before making it public for research is very important. Different privacy models include k-anonymity, l-diversity, t-closeness and differential privacy. This paper explores the role of quasi-identifiers and their roles for anonymization using differential privacy model. The research in this field can pave new ways for thinking before selecting quasi-identifiers for anonymization.