Anonymisation is the process by which sufficient elements of a dataset are removed to make identification impossible. It is a method of protecting privacy.
Anonymisation should not be confused with pseudonymisation. Unlike anonymisation, pseudonymisation is based on replacing directly identifying data with indirectly identifying data. This means that the identity of individuals can be recovered through the use of third-party data or encryption keys, whereas anonymisation is irreversible.
Since the anonymisation process seeks to eliminate any possibility of re-identification, several types of method will be used. However, these are all based on two principles:
- the transformation of data so that it no longer refers to a real person
- the generalization of data so that it is no longer specific to one person but common to a group of people.