Design and Development of Privacy Preservation Approach in Data Mining Using Noise Additive Model in Continuous and Multi-Dimensional Data Set

Data mining raises serious privacy issues. Privacypreserving data mining for multi-dimensional data sets, however, results in significant data loss, information leakage. An innovative method called Multidimensional Noise Additive Model (MNAM), which improves cluster identification and privacy while reducing information leakage, is suggested in this study. Anonymization technique that incorporates aggregation is applied in multidimensional data sets which leads to the reduction of information leakage and enhances the privacy. The domain-specific range of the subspaces are taken into consideration when performing noise addition. The MNAM method considers Euclidian distances among the neighboring clusters. Then, anonymized subspaces are subjected to random noise inside the subspace bounds to improve cluster detection and minimize data loss. Conventional algorithms fail to discover the types of clusters in datasets, MNAM is able to detect the majority of the original dataset clusters. Benchmark datasets are used to test MNAM, and the results demonstrate that MNAM can locate 85% of the underlying dataset clusters, in contrast to existing approaches that fail to locate any clusters. MNAM improves privacy by 65% while reducing data leakage by 60%.