🧬
Re‑identification
Anonymization
Process of linking de-identified data back to individuals via auxiliary data, linkage, or inference.
Definition
Re-identification occurs when de-identified or pseudonymous records are linked to a person using additional datasets, unique patterns, or identifiers. Risk depends on uniqueness, available auxiliary data, and attacker capabilities.
In plain English
Process of linking de-identified data back to individuals via auxiliary data, linkage, or inference.
Why this matters
Why it matters: Overstating “anonymous” claims can mislead users and increase privacy risk.
Example
Example: Assess re-identification risk before sharing data, minimize quasi-identifiers, and consider differential privacy for releases.