Question

A healthcare analytics team wants to share a de-identified BigQuery dataset with an external research partner. Before sharing, the compliance officer requires proof that individuals cannot be re-identified by combining quasi-identifiers such as ZIP code, birth date, and gender. Which Sensitive Data Protection (DLP) capability should the team use to measure this risk?

Accepted Answer

Run a k-anonymity risk analysis job on the quasi-identifier columns to compute equivalence class sizes

Answer

Run an inspection job with built-in infoTypes to detect PII in the columns

Answer

Apply a deterministic crypto-based tokenization de-identify transformation to the quasi-identifiers

Answer

Apply a date-shift transformation to the birth date column and redact the ZIP code