A Reliable Method for Accuracy Constrained Privacy Preservation for Relational Data

Today Information is most likely the vital and demanded resource. Distribution, transfer, mining and publishing data are primary operations in everyday life. Preserving the privacy of persons is indispensable one. Sensitive private information must be protected when data is published. The data should be uncovered through proper access control mechanism. Unauthorized data observation results in the disclosure of information to users not authorized to gain access to such information. There are two kinds of risks namely identity disclosure and attribute disclosure that affects privacy of persons whose data are published. Premature Researchers have contributed new methods specifically k-anonymity, l-diversity to protect privacy. K-anonymity method preserves privacy of persons against identity disclosure attack only. But Attribute disclosure attack makes negotiation this method. Limitation of k-anonymity is satisfied through l-diversity method. But it does not satisfy the privacy against the two attacks those are identity disclosure attack and attribute disclosure attack in several scenarios. To preserve the privacy of individuals sensitive information from attribute and identity disclosure attacks a new method is proposed. This method minimization information loss and gains the privacy by using generalization algorithm which is proposed in this method and is described in this paper.