Which technique adds randomness to the data analysis to protect privacy?

Which technique adds randomness to the data analysis to protect privacy?

• A. Masking
• B. Differential privacy
• C. Generalization
• D. Encryption

Answer: (B)

Differential privacy adds randomness to the results of data analyses to protect privacy. It does this by injecting calibrated noise into query outputs (or into the released data) so that the influence of any single individual’s data is limited. The amount of noise is tied to the query’s sensitivity and a privacy parameter, epsilon, which balances privacy and accuracy. The core idea is that whether a person’s data is included or not, the published results remain nearly the same, making it difficult to deduce information about that person. Masking hides actual values, generalization reduces precision, and encryption protects data in storage or transit but doesn’t provide the formal, randomized privacy guarantees across many analyses that differential privacy offers.

Differential privacy adds randomness to the results of data analyses to protect privacy. It does this by injecting calibrated noise into query outputs (or into the released data) so that the influence of any single individual’s data is limited. The amount of noise is tied to the query’s sensitivity and a privacy parameter, epsilon, which balances privacy and accuracy. The core idea is that whether a person’s data is included or not, the published results remain nearly the same, making it difficult to deduce information about that person.

Masking hides actual values, generalization reduces precision, and encryption protects data in storage or transit but doesn’t provide the formal, randomized privacy guarantees across many analyses that differential privacy offers.