Supermodularity Approach for Differential Data Privacy

Now a day the maximizing of data usage and minimizing privacy risk are two conflicting goals. The organization required set of transformation at the time of release data. While determining the best set of transformations has been the focus on the extensive work in the database community, the scalability and privacy are major problems while data transformation. Scalability and privacy risk of data anonymization can be addressed by using differential privacy. Differential privacy provides a theoretical formulation for privacy. A scalable algorithm is use to find the differential privacy when applying specific random sampling. The risk function can be employ through the supermodularity properties such as convex optimization.