Small sum privacy and large sum utility in data publishing.

Area of Science:

• Computer Science
• Data Privacy
• Information Security

Background:

• Existing privacy-preserving data publishing mechanisms have limitations in preventing all individual inferences.
• Recent research highlights that current methods may not fully address all inference attacks.

Purpose of the Study:

• To propose a novel model, SPLU (Sensitive data Protection with Limited Utility), that selectively counters inference attacks.
• To protect sensitive information (aggregate queries with small sums) while maximizing data utility for queries with large sums.

Main Methods:

• Developed the SPLU model, which differentiates between query types based on sum size.
• Introduced a sanitization algorithm within the SPLU framework to balance data protection and utility.
• Evaluated the model's effectiveness through empirical analysis.

Main Results:

• The SPLU model successfully protects sensitive information associated with aggregate queries yielding small sums.
• Queries with large sums receive higher accuracy, preserving data utility.
• Empirical results validate the model's intended behavior and effectiveness.

Conclusions:

• Not all inference attacks require the same level of mitigation; a selective approach can enhance data utility.
• The SPLU model offers a promising direction for privacy-preserving data publishing by optimizing for different query types.
• This approach maintains data utility for critical analytical queries while safeguarding sensitive individual data.