Primary source: EUR-Lex, (REGULATION (EU) 2016/679, 2016)
Secondary source: NIST
Anonymization is an irreversible process that removes the ability to identify the data subjects. For publication and storage of anonymized data issues of consent do not apply. Anonymization is not only important for Personal data (for example health data) but also for population units in other domains, for example, business data.
Publication of statistical tables requires usually Anonymization and other methods for respecting the legal rules for privacy of statistical information.
Anonymization is often not the preferred strategy for the use of Administrative data in the statistical production process because it devaluates the data for any further use. Instead Pseudonymisation is used.
Tools for Anonymization of Personal data are Data hashing techniques (one way hashing) and encryption techniques in which the decryption key has been discarded. In statistics frequently used techniques for statistical output are data swapping, random perturbation or generalisation (aggregation at a higher level).