Entity resolution · Privacy-preserving demonstration
Name Variant Matching Lab
Compare two names using S1 normalisation and matching. Inputs remain in your browser and are not submitted, stored or added to analytics.
Runs only in this browser · No request or analytics event
Candidate similarity
100.0%
- Normalised A
- mila kovalenko
- Normalised B
- mila kovalenko
- Token overlap
- 100.0%
- Edit similarity
- 100.0%
Identity verification required. A similarity score identifies a candidate for review. Confirm identifiers, dates, jurisdictions and primary records before deciding whether the names refer to the same subject.
What the score contains
The demonstration combines Unicode normalisation, Cyrillic and Arabic transliteration, diacritic folding, token overlap, edit similarity and a phonetic signal. It deliberately exposes the intermediate values so reviewers can challenge the result rather than treating one opaque number as a determination.