k-Anonymity (+ l-Diversity, t-Closeness)

← Back
Quasi-Identifiers (QIs)

Select columns that, when combined, could identify individuals (e.g., Age + City).

l-Diversity (intuitive): Each QI group must include at least l distinct values of a sensitive attribute. Prevents groups where everyone shares the same sensitive value (e.g., all “Cancer”).
t-Closeness (intuitive): Each QI group’s distribution for a sensitive attribute should be close to the overall dataset’s distribution. This demo uses Total Variation Distance and requires it ≤ t.
Sensitive Attribute Picker

Pick which attributes are sensitive. Use the filter to quickly find columns. Only picked ones appear below with settings.

Selected:
Selected Sensitive Attributes — Settings
Download Protected CSV