Privacy preserving data mining using statistical and data perturbation techniques

Abstract

The immense growth of technology in networking, storage and processing sectors has directed to the foundation for ultra-huge databases that store unprecedented amounts of information. The huge amount of information stored in the databases contains transactional data, multimedia data along with sensitive and private data of the customers or users. Numerous organizations and industries are in urge to explore the customer s private and sensitive information derived from data mining of massive repositories for analysis and to gain beneficial information. The mining of sensitive and private data of the public results in data misuse and leads to the privacy concern of an individual. Privacy has become the most significant problem in various data mining applications like healthcare, education, financial, sales and services. The healthcare data contains private information such as name, age, phone number and address. It also contains sensitive information like the name of the disease and nature of the disease. If this data gets into the hands of soiled third parties, they will misuse the data for their business benefit. So the data should be perturbed before it is released to the external parties. To overcome these privacy challenges, Privacy Preserving Data Mining (PPDM) is progressed as a solution. The objective of privacy preserving data mining is to secure sensitive and private information from being exposed to third-party vendors for analysis. newline