BASALT · JOURNAL
How to pseudonymize names consistently across PDFs
Pseudonymization preserves relationships in a document set—Participant 014 remains the same person everywhere—while separating the real identity into a restricted mapping. It reduces exposure but is not anonymity, because someone holding the key can reverse it.
Design the labels before editing
Use neutral identifiers that reveal nothing about the person. Avoid initials, birth years, department codes, or sequential labels assigned in a way that exposes enrollment order. Define how possessives, titles, family members, and organizations will be handled.
Keep the mapping table outside the PDF workspace with separate access controls. The released files should contain only the pseudonym, never comments like “replace Jane Smith with P014.”
Find identity variants across the set
Search full names, surname-first forms, initials, aliases, email usernames, signatures, and unique IDs. Review headers, filenames, bookmarks, form fields, and OCR layers. Then read for context that directly names the relationship between a pseudonym and real identity.
Straight text replacement can reflow lines or leave the old characters in the content stream. A safer PDF workflow redacts the original in place and overlays the neutral label, then verifies that the removed value no longer extracts.
Test consistency and separation
Search the output corpus for every real identifier and for each pseudonym. Confirm that one person never receives two labels and two people never share one. Inspect metadata, annotations, attachments, and old revisions for the mapping or original values.
Store the mapping, method version, and release hashes separately. If the key is destroyed, document that event because it changes what future researchers can do with the corpus.
Frequently asked questions
Is pseudonymized data anonymous?
No. A re-identification key exists, and contextual information may also reveal identity. Treat the output according to the governing risk and policy.
Can I use initials as pseudonyms?
Initials often remain identifying and may collide. Neutral generated labels are easier to control.
Why not use search and replace in a PDF?
Replacement can alter layout and may leave old objects recoverable. Redact the original content, add the label, and verify extraction.
Doing it in Basalt
Basalt removes original glyphs in place, preserves neighboring text, and verifies the absence of each identity value. Reproducible output and local processing support controlled research workflows without claiming to make the legal anonymization judgment. Get Basalt.
Redaction that proves itself
Basalt destroys the content you mark, then re-opens the file it wrote and proves the content is gone before it saves anything. Your documents never leave your Mac.