Abstract: Data mining is a process of extracting useful knowledge and hidden confidential patterns from data. The growth of information technology increases need for electronic data to be carefully stored and secretly shared. In last few decades a wide variety of approaches and techniques have been proposed for modifying data in such a way that privacy will remain preserved. Perturbation based randomization and SVD are one of the methods for preserving privacy. We need an algorithm which protects sensitive private information from huge databases in data mining system. In this paper we proposed an efficient privacy preserving data mining technique using perturbation based randomization in combination with SVD. In this technique we will apply several classification schemes on perturbed data. Experimental comparisons will define the effectiveness of this algorithm.

Keywords: Randomization; perturbation; SVD.