Data Distortion Approaches for Privacy Preserving in Association Rule Mining
B. Janakiramaiah, RamaMohan Reddy A, Gunjan Kalyani · International Journal of Advances in Computing and Information Technology · 2012
Huge repositories of data have sensitive information that must be confined against unauthorized access. The safeguard of the privacy of this data has been a long-standing goal for the database security research community. Current advances in data mining algorithms have enhanced the disclosure risks that one may come across when releasing data to remote parties. To address this challenging problem, different data distortion approaches were projected to protect sensitive data in transactional databases. We introduce taxonomy of sanitizing algorithms and validate all distortion algorithms against real and synthetic data sets. A set of metrics are considered to assess the effectiveness of the algorithms in terms of information loss and to quantify how much confidential information has been disclosed. We also perform experimental study for evaluation and comparison of different distortion algorithms.