Privacy Evaluation: Systematic Approach

This lecture covers a systematic approach to privacy evaluation, focusing on concepts like confidentiality, pseudonymity, anonymity, unlinkability, and unobservability. It explains how to model privacy mechanisms, determine what adversaries can observe, and evaluate privacy properties. The case study on SwissCovid illustrates practical applications of these concepts, emphasizing the importance of achieving anonymity and confidentiality in sensitive data sharing. Strategies to enhance privacy, such as obfuscation and plausible deniability, are discussed in detail, along with the role of probabilistic analysis. The lecture concludes with insights on common techniques, adversary models, and the collaboration between privacy and security researchers.