Enhancing Utility and Privacy using t-closeness For Multiple Sensitive Attributes

Organizations publish the individual’s information in order to utilize the data for the research purpose. But the confidential information about the individual is revealed by the adversary by combining the various releases of the several organizations. This is called as linkage attacks. This attack can be avoided by the SLOMS method which vertically partitions the single quasi table and multiple sensitive tables. The SLOMS method uses MSB-KACA algorithm to generalize the quasi identifier table in order to implement k-Anonymity and bucketizes the sensitive attribute table to implement l-diversity. But there is a chance of probabilistic inference attack due to bucketization. So, the method called t-closeness can be applied over MSB-KACA algorithm which compute the value using Earth Mover Distance(EMD) and set the minimum value as threshold in order to equally distribute the attributes in the table based on the threshold ’t’. Such that the probabilistic inference attack can be avoided. The performance of t-closeness gets improved and evaluated by Disclosure rate which becomes minimal while comparing with MSB-KACA algorithm.