Hiding Predictive Association Rules on Horizontally Distributed Data

Abstract. In this work, we propose two approaches of hiding predictive association rules where the data sets are horizontally distributed and owned by collaborative but non-trusting parties. In particular, algorithms to hide the Collaborative Recommendation Association Rules (CRAR) and to merge the (sanitized) data sets are introduced. Performance and various side effects of the proposed approaches are analyzed numerically. Comparisons of non-trusting and trusting third-party approach are reported. Numerical results show that the non-trusting third-party approach has better processing time, with similar side effects to the trusting third-party approach.

• Type: Journal
• Year: 2009
• Where: IEAAIE
• Authors: Shyue-Liang Wang, Ting-Zheng Lai, Tzung-Pei Hong, Yu-Lung Wu
• Keywords: Artificial Intelligence, Data Sets, IEAAIE 2009, Third-party Approach, Trusting Third-party Approach