Trust
De-identification and anonymisation
Removing names is rarely enough to make data anonymous. De-identification is a set of techniques that reduce the risk that individuals can be re-identified.
Pseudonymised vs anonymised
Pseudonymised data replaces direct identifiers with codes, but can still be linked back with extra information. Under the GDPR it is still personal data. Anonymised data cannot reasonably be linked to a person by any means likely to be used, and falls outside the GDPR.
Common techniques
| Technique | Idea |
|---|---|
| Aggregation | Publish counts or averages instead of records. |
| Generalisation | Coarsen values (age bands, postcode districts). |
| Suppression | Remove rare values or small cells. |
| k-anonymity | Each record is indistinguishable from at least k−1 others on key attributes. |
| Differential privacy | Add calibrated noise so results barely depend on any one person. |
| Synthetic data | Generate artificial records that mimic statistical properties. |
Work with ex omnia
Contact details for ex omnia will be published here once confirmed.