Privacy Preserving Data using Fuzzy Hybrid Data Transformation Technique

Objective: To provide improved privacy for dataset using fuzzy set properties with optimal Information loss, thereby achieving better utility. Methods: We propose Fuzzy Hybrid data Transformation method by combining fuzzy data modification method and Random Rotation Perturbation techniques (RRP). Fuzzy data modification method contains fuzzy K-member clustering along with membership function to be executed to distorted data. RRP preserves the geometric structure on dataset. Findings: Experimentation proves that our method gives least information loss when compared with existing method along with fuzzy membership function individually for different values of k. Thereby, we achieve dual goal of privacy and utility. Applications: The experiment has been done over the Adult dataset derived from UCL Repository and used for numerous applications such as analysis, mining, forecasting and prediction etc.