A Novel Anti-data Mining Technique Based On Hierarchical Anti-clustering (HAC)

Data mining is a technique to search potential valuable information from databases. Preventing personal data and high security data therefore pose a difficult task to IT experts. In this paper, we propose a novel anti-data mining (ADM) database security scheme, that protect against data mining. The scheme makes use of hierarchical clustering where noise is added to change the cluster structure of data. The proposed hierarchical anti-clustering (HAC) scheme modifies the cluster structure of the original data. Experimented results show that data may be protected against during the HAC key can be used reverse the cluster structure to its original. At the meantime, HAC also designs the key value to restore correctly the protected database.