Name matching and transliteration tool
Try a Preset
Direct Match
89.9%
High Match
After Transliteration
89.9%
High Match
Best Permutation
94.7%
High Match
Processing steps
Transliterated Input
Mohammed Al-Rahman
Transliterated Target
Mohammad Al Rahman
Name Permutations Tested
Best Match Pair
How fuzzy name matching works in sanctions screening
Effective fuzzy name matching for sanctions screening involves three key stages:
1. Transliteration: Converting names from non-Latin scripts (Arabic, Cyrillic, Chinese) into standardised Latin characters. Multiple transliteration standards exist — for example, the Arabic name محمد can be romanised as Mohammed, Muhammad, Mohamed, or Muhammed.
2. Name Permutation: Testing all reasonable orderings of name components. Different cultures place family names in different positions, and middle names may or may not be included in watchlist entries.
3. Similarity Scoring: Algorithms like Jaro-Winkler measure how closely two strings match, accounting for transpositions, insertions, and deletions. FirmScreen uses a weighted scoring model that combines name similarity with additional fields like date of birth and nationality.
FAQ
Questions firms ask
Common questions about name matching in sanctions screening.
Still have a question?
hello@firmscreen.io
Names are rarely entered identically across databases. Transliterations from non-Latin scripts (Arabic, Cyrillic, Chinese) can produce multiple valid spellings. Cultural naming conventions vary — some cultures place family names first, others use patronymics. Sanctions lists may use different romanisation standards than your client records. Fuzzy matching algorithms must handle all these variations while minimising false positives.
Jaro-Winkler is a string similarity metric that measures the edit distance between two strings, giving extra weight to matching prefixes. It returns a score between 0 (no similarity) and 1 (exact match). It is widely used in sanctions screening because it handles common name variations like transpositions, missing characters, and spelling differences well. This tool uses Jaro-Winkler to illustrate the idea. FirmScreen’s own matching is deterministic and also uses aliases, transliteration and date of birth.
Transliteration is the process of converting text from one writing system to another. For example, the Russian name "Владимир" can be transliterated as "Vladimir", "Wladimir", or "Vladimeer" depending on the romanisation standard used. Effective sanctions screening must handle these variations to avoid missing true matches.
Production-grade fuzzy matching
FirmScreen's matching engine goes beyond this demo with configurable field weights, alias comparison, and multi-list screening in a single operation.
Book a demo