Anonymizing 1:M microdata with high utility.

Area of Science:

• Computer Science
• Data Privacy
• Information Security

Background:

• Data publishing and mining require robust privacy preservation.
• Existing privacy models often fail with datasets containing multiple records per individual (1:M datasets), leading to new privacy risks.

Purpose of the Study:

• To address disclosure risks in 1:M data publishing.
• To propose a novel privacy model and an efficient algorithm for privacy preservation and data utility.

Main Methods:

• Introduction of the (k, l)-diversity privacy model.
• Development of the 1:M-Generalization algorithm for 1:M datasets.
• Comparison of the proposed algorithm with alternative approaches using real-world data.

Main Results:

• The proposed (k, l)-diversity model effectively handles privacy risks in 1:M datasets.
• The 1:M-Generalization algorithm demonstrates superior performance compared to state-of-the-art techniques.
• Experiments show improved data utility and reduced computational cost.

Conclusions:

• The (k, l)-diversity model and 1:M-Generalization algorithm offer a significant advancement in privacy-preserving data publishing for 1:M datasets.
• The approach balances privacy preservation with data utility more effectively than existing methods.