Privacy Preserving Data Mining: Techniques, Classification and Implications - A Survey

Abstract

Privacy has become crucial in knowledge based applications. Proper integration of individual privacy is essential for data mining operations. This privacy based data mining is important for sectors like Healthcare, Pharmaceuticals, Research, and Security Service Providers, to name a few. The main categorization of Privacy Preserving Data Mining (PPDM) techniques falls into Perturbation, Secure Sum Computations and Cryptographic based techniques. There exist tradeoffs between privacy preservation and information loss for generalized solutions. The authors of the paper present an extensive survey of PPDM techniques, their classification and give a preliminary implication of technique to be used under specific scenarios.