Efficient Table Anonymization for Aggregate Query Answering
Source: AT&T Intellectual Property
Privacy protection is a major concern when microdata need to be released for ad hoc analyses. This has led to a lot of recent research in privacy goals and table anonymization schemes, such as k-anonymity, l-diversity, t-closeness and (k, e)-anonymity. It is important that the table anonymization preserves sufficient information to support ad hoc aggregate queries over the data, and to return reasonably accurate answers. The recent framework of permutation-based anonymization was shown to be better than generalization based approaches in answering aggregate queries with sufficient accuracy, while providing strong anonymity guarantees. This paper focuses attention on the case when the sensitive attributes are numerical (e.g., salary) and (k, e)-anonymity has been shown to be an appropriate privacy goal.