Automated zone design as an aid to effective anonymisation

Robards, James, Martin, David and Gale, Chris (2016) Automated zone design as an aid to effective anonymisation. New Approaches to Anonymisation, Cambridge, United Kingdom.

Abstract

Geographical aggregation is a long-established tool in the effective anonymisation of personal data records. This applies both to the preparation of national datasets for open publication, such as small area census statistics, and to the protection of microdata, such as the limitation on geographical coding in the ONS Longitudinal Study to Government Office Region level. Choice of geographical aggregation scale is likely to interact with access controls and other aspects of the analysis environment (e.g. highly controlled access environments for microdata). A key risk with individual records is that an intruder may be able to use external data relating to the same geographical units in order to identify (find someone) or attribute characteristics (learn new information about someone). A central challenge therefore is the determination of aggregation levels which provide sufficient protection without excessively affecting utility from the users’ perspective. This presentation considers the contribution of automated zone design to effective anonymization, ranging from its use in 2011 Census data production by the Office for National Statistics (ONS) to current work at the Economic and Social Research Council (ESRC) funded National Centre for Research Methods (NCRM) using synthetic data to explore bespoke aggregation scenarios for use in protection of linked administrative data

This record has no associated files available for download.

Contributors

Author: James Robards

Author: David Martin

Download statistics