How Appropriate is K-Anonymity for Addressing the Conflict Between Privacy and Information Utility in Microdata ASnonymisation

Before statistical data, such as microdata, can be released to the public, it needs to be anonymised. Anonymisation protects the privacy of the individuals whose data is released. However, as microdata is anonymised, its level of privacy increases, while its level of information utility decreases. K-anonymity is often used to address the conflict between privacy and information utility in microdata anonymisation. In this paper, we determine the extent to which k-anonymity is appropriate for addressing this conflict. We argue that the way in which k-anonymity is currently used to address this conflict does not necessarily lead to an optimum balance between privacy and information utility. We also provide recommendations for an appropriate solution for addressing the conflict between privacy and information utility.