Towards Balancing Data Usefulness and Privacy Protection in K-Anonymisation

K-anonymisation, as an approach to protecting data privacy, has received much recent attention from the database research community. Given a single table, there can be many ways to anonymise it. So criteria for determining a preferred solution is important. Various techniques have been proposed, all attempting to achieve some form of optimality in k-anonymisation, but few have considered the balance between the usefulness of the anonymised data and the protection of the original. In this paper, we address this issue and propose a two-step approach which allows data usefulness and privacy protection requirements to be considered and balanced in k-anonymisation.