Optimizing Privacy-Preserving Data Mining Model in Multivariate Datasets
Sharath Yaji, Neelima Bayyapu · 2019
The increased importance of data protection for sensitive or private data paved a new research direction towards privacy-preserving in data mining. This article shares work-in-progress research data privacy-preserving research carried out by the authors. This work focuses on i) The study of different k-anonymization algorithms, ii) Classifying sensitive and non-sensitive data, iii) Protection model against differential privacy attacks. Our observations show i) Depending on the size of the dataset, the performance of the k- anonymization algorithms varies for sequential and parallel computations. ii) The sensitive and nonsensitive data can be differentiated through correlation coefficients. iii) Differential privacy application attacks can be avoided by replacing existing Laplacian with other partial differential equations as noise.