Overview on K-anonymity Model for Overlapped Attributes

Abstract

K-anonymity model is mostly used technique of privacy preserving data publishing. In K-anonymity model data is converted into anonymous state. So, that adversary can’t be able to disclose sensitive information about the user. Generalization and suppression are most commonly used anonymity technique, but generalization contains some drawbacks i.e. generalization disturbs correlations between attributes. In this paper a novel model is proposed which uses generalization technique specially to maintain correlation among overlapped attributes and way to reduce dimensionality of data set. Experimental evaluation section shows efficiency and correctness of proposed model.