Question

A public health agency wants to release a data set of hospital admissions to external researchers. The raw data includes ZIP code, date of birth, sex, and diagnosis. The privacy engineer must ensure no individual can be singled out by combining the quasi-identifiers, while keeping the data useful for demographic trend analysis. Which technique should the engineer apply to the quasi-identifiers before release?

Accepted Answer

Generalize ZIP code to the first three digits and date of birth to year, then verify each combination of quasi-identifiers maps to at least k records

Answer

Encrypt the ZIP code and date of birth fields with AES-256 and share the decryption key with the researchers

Answer

Hash the diagnosis field with SHA-256 so the sensitive attribute cannot be read directly

Answer

Replace each patient's ZIP code with a random token stored in a reversible lookup vault