Records / De-identification rules
Several kinds of evidence, never one detector
Four layers: the people your exports list in their own structure (senders, authors, users, contacts), found wherever they are mentioned; patterns checked against each identifier's own rule (check digits, key formats); names from public dictionaries, greetings, titles and sign-offs; and your own lists of words to remove and to keep.
Dated 1 Jul 2022Source checked 15 Oct 2026Page out of reach at the last check, 16 Oct 2026Basis: our standard
“rule-based systems excelled at finding certain kinds of information (e.g., SSNs and phone numbers), while the trainable systems worked better for other kinds of data”
NIST IR 8053, De-Identification of Personal Information, p. 31, checked Oct 2015
Everything read from this source (3)Archived copy, 10 Oct 2026
Pilán et al., The Text Anonymization Benchmark (TAB), Computational Linguistics, checked 1 Jul 2022
HHS, Guidance on de-identification of protected health information, checked 26 Nov 2012
Everything read from this source (5)Archived copy, 5 Oct 2026
- Stand-ins from a secret key you keep, never a plain hash
- A stand-in stays the same within a file, unless the use needs more
- Every mention of a person, or the person is not protected
- The engine is measured on a public benchmark
- Details that identify together are coarsened and measured
- Content is screened as well as names
It is shown on its own, not counted into a figure.
Cited in this guide