What is fuzzy matching?
Fuzzy matching in AML screening scores how similar an input name is to a listed name so “Mohammed Al Saud” and “Mohamed Alsoud” can still surface as a possible match.
Why exact match fails
Arabic, Latin, and mixed scripts, honorifics, and data-entry typos mean exact equality misses real risk. Exact match also explodes false positives if you lowercase and strip spaces without a review queue.
How to operate it
Send the fullest legal name you have, plus date of birth or nationality when available. Route possible matches to REVIEW. Confirm or clear with the screening id on the case. Do not auto-BLOCK every similar string.
Naiza’s role
Naiza returns structured match context for documented watchlists. Tune operations around REVIEW volume, not around inventing an unpublished matching algorithm. Confirm current fields in docs.
Frequently asked questions
Short answers written so search and answer engines can cite them.
What is fuzzy name matching?
It is a comparison that allows small differences in spelling or word order so watchlist screening can surface possible matches instead of only exact strings.
Does fuzzy matching increase false positives?
It can. That is why mature programs use REVIEW, extra identity fields, and analyst confirmation instead of blocking every near-match.
Should we disable fuzzy matching to reduce queue size?
Only with a documented risk acceptance. Exact-only screening is easier to operate and easier for an evader to bypass with a one-letter change.