Hiding critical transactions using a modified un-realisation approach

Huge amount of data generated from the hospitals, social media sites and online departmental stores contains sensitive data. Nowadays finding association rules among these data may lead to privacy issues, in turn to leakage of sensitive information that may reveal the facts about an individual. In this paper an algorithm is proposed for hiding the sensitive association rules by minimising the ghost rules and lost rules. This maximum association rule (MAR) hiding performs better than GA and PSO algorithms. These traditional algorithms may lack in hiding the association rules due to its computational complexity, hiding failure, lost rules and ghost rules. The performance of this approach has been improved by using a modified un-realisation algorithm for hiding the sensitive association rules.