Anonymization Techniques for Privacy Preservation of published Data

Vidya Ingle

In today’s world data is collected for various purposes. The collected data includes personal details or some other confidential information. Database is a collection of data that can be accessed, updated and it enables user to retrieve data. Providing security to these databases is big issue now days. Today online information is stored for almost everything like-Online admission, online banking, online shopping, E-Health service and social networking etc. This information can be used by insurance companies, drug manufacturing companies and various other marketing agencies without consent of the data provider. To maintain confidentiality, unauthorized third parties must be prevented from accessing and viewing data. It is also essential to maintain database integrity while data is transferring from source to destination. Suppose the data owner wants to share the data with researchers or analysts, how can a data owner technically insure that the individuals who are the subjects of the data cannot be re- identified while the data remain practically useful? Various anonymization techniques, like suppression, generalization and bucketization are designed for preserving privacy of published data. Generalization loses significant amount of information especially for large volume of data and bucketization is not sufficient prevent membership disclosure and cannot be used for data where quasi-identifying attributes and sensitive attributes cannot be clearly separated. . Slicing is new privacy preserving technique which provides improved data utility than generalization and is more efficient in workload pertaining to the sensitive attribute