Performance Analysis of Various Parameters in Sensitive Association Rule Hiding For Privacy in Distributed Collaborative Data Mining

Abstract

Privacy preserving or Data and knowledge hiding is a novel research area in distributed collaborative data mining to protect the privacy of confidential or sensitive information of individuals. Many of the researchers have been proposed methods in Privacy Preserving Data mining (PPDM) to hide sensitive information in association rule mining. Association rule hiding is the process to modify the original database for vanishing sensitive association rule while generating rules using rule mining algorithms. The better rule hiding methods are not affecting the quality of the database and non-sensitive rules. In this paper, privacy preserving association rule hiding methods in the literature are studied in detail to find the problem in each method and metrics used for evaluating these methods. The performance of metrics, merit and demerit of every method are thoroughly compared. Finally, the remarkable future direction is suggested in the association rule hiding area based on the problems has been found from the literature.