Which technique ensures that a dataset cannot be distinguished from at least k-1 other records?

Which technique ensures that a dataset cannot be distinguished from at least k-1 other records?

• A. K-anonymity
• B. Differential privacy
• C. Pseudonymization
• D. Aggregation

Answer: (A)

K-anonymity ensures that a dataset cannot be distinguished from at least k-1 other records based on the attributes used to identify individuals. In practice, we look at quasi-identifiers—attributes like age, ZIP code, or gender that, alone or together, could help pinpoint someone if combined with outside information. By generalizing or suppressing these attributes, the dataset is transformed so that every combination of quasi-identifiers appears in at least k records. That means any given record shares its identifying profile with at least k−1 others, making it impossible to single out a unique individual from the released data. Differential privacy adds random noise to query results to limit what can be learned about any one person, rather than guaranteeing indistinguishability within the released data itself. Pseudonymization replaces direct identifiers with aliases but can still leave other attributes that could re-identify someone when linked with external data. Aggregation reduces detail but doesn’t ensure that every record sits in a group of size at least k based on the identifying attributes. So the approach that achieves the stated goal is k-anonymity.

K-anonymity ensures that a dataset cannot be distinguished from at least k-1 other records based on the attributes used to identify individuals. In practice, we look at quasi-identifiers—attributes like age, ZIP code, or gender that, alone or together, could help pinpoint someone if combined with outside information. By generalizing or suppressing these attributes, the dataset is transformed so that every combination of quasi-identifiers appears in at least k records. That means any given record shares its identifying profile with at least k−1 others, making it impossible to single out a unique individual from the released data. Differential privacy adds random noise to query results to limit what can be learned about any one person, rather than guaranteeing indistinguishability within the released data itself. Pseudonymization replaces direct identifiers with aliases but can still leave other attributes that could re-identify someone when linked with external data. Aggregation reduces detail but doesn’t ensure that every record sits in a group of size at least k based on the identifying attributes. So the approach that achieves the stated goal is k-anonymity.