Managing privacy of sensitive attributes using fuzzy-based data transformation methods in privacy preserving data mining environment

Abstract

When we extract personal, sensitive and business information in data mining applications, then certain problems occurs. Privacy attack occurs due to the misuse of individual information. In centralised database environment, data transformation methods in fuzzy-based data in the field of privacy preserving clustering are proposed in this paper. In first case, a fuzzy data transformation method is proposed and different experiments are conducted by changing the fuzzy membership functions such as Z-shaped fuzzy membership function, Triangular fuzzy membership function, Gaussian fuzzy membership function to transform the original dataset. In second case, a hybrid method is proposed as a combination of fuzzy data transformation approach which is specified in first case and random rotation perturbation (RRP). The experimental outcome verified that the hybrid approach permits finest results for every member functions.