Measuring Privacy: Critical Reflections and Directions for a Metrology-Based Approach
Abstract
Privacy measurement in digital systems lacks a standardised metrological framework to ensure reliable and comparable assessments. A metrological approach ensures that privacy measurements are reliable, reproducible, and comparable over time and across different contexts. In this work most common privacy metrics, including k-anonymity, ℓ-diversity, t-closeness, differential privacy, and mutual information, are critically evaluated, identifying their strengths and limitations from a metrological perspective. Initial directions and open challenges toward a metrology-based approach to privacy measurement are outlined, too.
Event details
- Event
- TC6 M4Dconf2025
- Technical Committee
- TC6
- info@m4dconf.org
- Place
- Benevento, ITALY
- Time
- 3 September 2025 - 5 September 2025
- Website
- https://www.m4dconf.org/