Differential Inference Testing A Practical Approach to Evaluate Anonymized Data

Ali Kassem (1) , Gergely Acs (2) , Claude Castelluccia (3) , Catuscia Palamidessi (4)

1 COMETE - Concurrency, Mobility and Transactions
2 BME - Budapest University of Technology and Economics [Budapest]
3 PRIVATICS - Privacy Models, Architectures and Tools for the Information Society
4 MSR - INRIA - Microsoft Research - Inria Joint Centre

In order to protect individuals' privacy, data have to be ``well-sanitized'' ( ``well-anonymized'') before sharing them, i.e. one has to remove any personal information before sharing data. However, it is not always clear when data shall be deemed well-sanitized. In this paper, we argue that the evaluation of sanitized data should be based on whether the data allows the inference of sensitive information that is specific to an individual, instead of being centered around the concept of re-identification. We propose a framework to evaluate the effectiveness of different sanitization techniques on a given dataset by measuring how much an individual's record from the sanitized dataset influences the inference of his/her own sensitive attribute. Our intent is not to accurately predict any sensitive attribute but rather to measure the impact of a single record on the inference of sensitive information. We demonstrate our approach by sanitizing two real datasets in different privacy models (k-anonymity, l-diversity, and differential privacy) and evaluate/compare each sanitized dataset in our framework.

Mots clés: Inferences; Machine Learning; l-diversity; k-anonymity; Anonymization; Differential Privacy; Sanitization

Domaines: Cryptographie et sécurité [cs.CR]

Ali Kassem, Gergely Acs, Claude Castelluccia, Catuscia Palamidessi. Differential Inference Testing A Practical Approach to Evaluate Anonymized Data. [Research Report] INRIA. 2018, pp.1-21. ⟨hal-01681014v2⟩