K-Anonymity

Noun · Security & Infosec

Definitions

  1. A privacy property in which each record in a released dataset is indistinguishable from at least k-1 other records with respect to a chosen set of identifying attributes. K-anonymity reduces straightforward re-identification risk, but it can still fail if the sensitive values within each group are too revealing.

    In plain English: A privacy rule that makes each person in a dataset look like at least several others based on identifying details.

    Example: "The dataset met k-anonymity on paper, but attackers could still infer health conditions from the lack of diversity inside each group."

Related Terms