🎭
l‑Diversity
Anonymization
Extension of k-anonymity requiring diversity of sensitive attributes within each equivalence class.
Definition
l‑diversity requires that each k‑anonymous equivalence class contains at least l “well-represented” values of a sensitive attribute. It helps prevent homogeneity attacks but can still fail with skewed distributions.
In plain English
Extension of k-anonymity requiring diversity of sensitive attributes within each equivalence class.
Why this matters
Why it matters: It reduces attribute disclosure beyond simple re-identification resistance.
Example
Example: Ensure each generalized group contains multiple diagnoses, not a single dominant diagnosis.