International Journal of Computing Science and Information Technology (IJCSIT)
Managing a data set with sensitive but useful information, such as medical records, requires reconciling two objectives: providing utility to others and respecting the privacy of individuals who contribute to the data set. The earliest privacy preserving technique was encryption. Following it is the k-anonymity where each tuple in a k-anonymized dataset should appear at least k times. A series of recent papers formalizes the notion of differential privacy. A database privatization mechanism satisfies differential privacy if the addition or removal of a single database element does not change the probability of any outcome.