Maintaining K-Anonymity on Real-Time Data
Gary Blosser, Justin Zhan · 2007
The usage of K-anonymity to protect static data sets is well known, but when applied to real time data personal privacy may be breached. For example, a hospital that releases information on all current patients may begin to involuntarily disclose private information due to the presence of a long-term patient records which, when identified through time, reveal identifying information about other records contained in the same k-anonymous tuples. In this paper, we will give a feasible scenario, brief overview of the K-anonymization method, the flaws arising in real-time data, and a practical solution to counter the problems. The basic premise of the solution is to track the released k-anonymized tuples and, in the future, prevent the release of the same tuple at a decreased privacy level.