Heuristic approach for association rule hiding using ECLAT
Melissa Femandes, Joanne Gomes · 2017
In many organizations huge amount of data is generated. Organizations use this data for their own benefit. Data mining extracts useful knowledge from huge data. Association rule mining is a powerful technique to find hidden patterns in large database. The limitation of mining association rules is that some sensitive patterns are revealed from sensitive rules. It is necessary to hide sensitive rules due to privacy concerns. The technique used to hide sensitive rules in order to get sanitized data is called Association rule hiding. The paper aims at proposing a methodology which includes blend of equivalent class transformation (ECLAT) algorithm used to find frequent item sets and heuristic approach which is one of the methods of association rule hiding. The proposed hybrid approach is evaluated on the basis of execution time, dissimilarity, lost rules, hiding failure and number of transactions modified.