3 ms·
K-Anonymity
- prepend 6y agoIt’s interesting how this is such a useful technique as part of de-identifying datasets but there aren’t many packages for calculating optimal or even simple changes for data to make them k-anonymous. Wikipedia calls out k-optimize but the paper just has pseudo-code and I expected more reusable packages to exist. I’ve used sdcmicro in R [0], but it isn’t practical for large datasets with multiple quasi-identifiers. [0] http://www.ihsn.org/software/disclosure-control-toolbox http://www.ihsn.org/software/disclosure-control-toolbox