Page 751

Ming Ding | Information Security and

Privacy Group | Data61, CSIROPrivacy risk assessment –Examples

•Attribute inference: 𝐼𝑠;𝑦

➢Intuition: The amount of sensitive information of 𝑥that can be inferred by the observation 𝑦

➢This metric is based on the mutual information shared between the sensitive information 𝑠and the

observation 𝑦

➢This metric can be evaluated using Information theory or empirical experiments

➢The smaller the metric, the stronger the privacy protection

Original data x

(sensitive info: s

non-sensitive info: u)Output/Observation y

(data, query answer, AI/ML

models/parameters, etc.)