Preserving the Privacy and Sharing the Data using Classification on Perturbed Data

Data mining is a powerful tool which supports automatic extraction of unknown patterns from large amounts of data. The knowledge extracted by data mining process support a variety of domains like marketing, weather forecasting, and medical diagnosis .The process of data mining requires a large data to be collected from diverse sites. With the rapid growth of the Internet, networking, hardware and software technology there is tremendous growth in the amount of data collection and data sharing. Huge volumes of detailed data are regularly collected from organizations and such datasets also contain personal as well as sensitive data about individuals. Though the data mining operation extracts useful knowledge to support variety of domains but access to personal data poses a threat to individual privacy. There is increased concern on how sensitive and private information can be protected while performing data mining operation. Privacy preserving data mining algorithms gives solution for the privacy problem. PPDM gives valid data mining results and also guarantees privacy protection for sensitive data stored in the data warehouse. In this paper we analyzed the threats to privacy that can occur due to data mining process. We have proposed a framework that allows systemic transformation of original data using randomized data perturbation technique and the modified data is submitted as a result of query to the parties using decision tree approach. This approach gives the valid results for analysis purpose but the actual or true data is not revealed and the privacy is preserved.