Hybrid Algorithm for Privacy Preserving Association Rule Mining | Science Publications

Journal of Computer Science

Ila Chandrakar¹  Kondabala Renuka¹  Mortha Manasa¹  Yelipe Usha Rani¹ 

关键词: Hybrid algorithm;    association rule mining;    privacy preserving;    mining algorithms;    sensitive items;    hiding association;   

DOI  :  10.3844/jcssp.2010.1494.1498

学科分类：计算机科学（综合）

来源: Science Publications

【 摘 要 】

Problem statement: The objective of the hybrid algorithm for privacy preserving datamining is to hide certain sensitive information so that they cannot be discovered through associationrule mining techniques. Approach: The sensitive items whether in Left Hand Side (LHS) or RightHand Side (RHS) of the rule cannot be inferred through association rule mining algorithms bycombining the concept of Increase Support of Left Hand Side (ISL) and Decrease Support of RightHand Side (DSR) algorithms i.e., by increasing and decreasing the support of the LHS and RHS itemof the rule respectively. Results: The efficiency of the proposed approach is compared with aloneIncrease Support of Left Hand Side (ISL) approach for real databases on the basis of number of rulespruned. Conclusion: The hybrid approach of ISL and DSR algorithms prunes more number ofsensitive rules with same number of database scans.

【 授权许可】

Unknown