Special Issue Journal of Human Geography and Environmental Studies

Special Issue 1, December 2009

Article Number: 10

Download [PDF] 

Association rules for mining in vertically and horizontally partitioned databases while maintaining privacy of data

M.M. Madani Lemraski and H. Asefi

Pages 52-57

Abstract—The main objective of Data Mining is generalization of information and not to popularization of private data. We can implement Data Mining and maintain the privacy preserving capability of data simultaneously. Data Mining can deal with private data privately. Ergo, the true problem is not data mining, but the way to do it. This paper introduces two algorithms for exploring frequentative Itemsets by maintaining the privacy of data owners: one algorithm is for vertically partitioned data and another for horizontally partitioned data. Data owners will collaborate with each other based on the introduced algorithms, and after exploring repetitive Itemsets, they will develop association rules without exposing private data. We have presented these algorithms for multi party state. We have then analyzed privacy preserved and communication and complexity. These algorithms then have been analyzed with other algorithms from accuracy and preserving privacy point of the view.

Keywords—associations rule mining, data perturbation, horizontally partitioned data, vertically partitioned data.