A SURVEY OF HIDING ASSOCIATION RULE METHODS IN TRANSACTION DATASETS

Abstract

Privacy-Preserving Data Mining (PPDM) is a relatively new area of ​​research in the data mining community and has been studied for over a decade. PPDM researches the side effects of data mining methods that stem from intrusions into the privacy of individuals and organizations. Some approaches to solving this problem have been investigated and applied. The proposed methods can group into two main directions: data hiding and knowledge hiding. Data hiding is a research direction on the privacy of raw data or information protected during data mining. The methods of this group work on the data itself to hide sensitive information by different algorithms. Knowledge hiding involves protecting the results of sensitive data mining, not the raw data itself. This is the main application direction of data mining tools and algorithms. In which, association rule hidden is a research direction in knowledge hidden group. In this paper, we focus on investigating different association rule hiding techniques and evaluating the effectiveness of the proposed approaches.