Credit decision dataset de-identification is the removal of applicant fields from a decision table. ECOA / Reg B (12 CFR 1002) governs fair treatment in such credit. anonym.plus marks each item on your device, so the analytic signals stay usable while direct identifiers go.
When this applies
A decision table ties applicants to scores, outcomes, and protected-class proxies. You strip direct fields before the data trains a model or feeds a fair-lending review.
How anonym.plus handles it
- Open the decision table in anonym.plus on your device.
- The tool flags names, SSNs, and contact fields.
- Local OCR reads any scanned supporting pages.
- Keep the score and outcome columns you must analyze.
- Swap or black out the confirmed direct fields.
- Save the clean dataset locally.
What you need to provide
- The decision table (CSV, XLSX, or PDF).
- An operator (Replace keeps the columns readable).
- Optional allow-list for score and outcome columns.
PII & financial identifiers detected
| Category | anonym.plus entity type | Example |
|---|---|---|
| Names | PERSON | applicant id → [APPLICANT] |
| Identifiers | US_SSN | 455-21-7763 → [SSN] |
| National ID | NATIONAL_ID | ITIN row → [TAXID] |
| Money | MONEY | income field → [AMOUNT] |
| Dates | DATE_TIME | decision date → [DATE] |
| Contact | EMAIL_ADDRESS | contact email → [EMAIL] |
Compliance achieved
- Supports fair-lending analysis under ECOA / Reg B (12 CFR 1002).
- Keeps the score and outcome signals you must analyze.
- Offline work keeps the dataset off any server.
Anonymize credit decision datasets offline — see plans & start free →
Limitations & cautions
Direct identifiers are not the only risk. Score, income, and date combine into quasi identifiers that can re-link a person. Generalize those for true de-identification, and keep the name map turned off.
Frequently asked questions
Is removing the applicant id enough for anonymity?
No. Score, income, and date can re-link a person when combined. Generalize or suppress those quasi identifiers for stronger de-identification.
Can I keep the outcome columns for analysis?
Yes. Allow-list the score and outcome fields so they stay while direct identifiers are removed.
Is the dataset uploaded for processing?
No. The app runs locally. For real anonymity, keep the reversible name map turned off.