TLDR: Fuzzy matching is used to identify similar but not exactly matching names, helping to avoid missed matches in PEP and Sanctions screening. It employs the Levenshtein Distance algorithm, allowing up to one character difference to reduce false positives while capturing spelling errors or variations.
Fuzzy Matching
Fuzzy matching is the process used to match names that appear similar but may not be spelt exactly the same. This is standard across the industry and is used to avoid unnecessary failures if the names are mathematically close enough.
Misspelt names are most likely to be an error on the client's part. However, using slightly different names across different applications or records could be an indicator of risk.
This is why we employ fuzzy matching in our PEP and Sanctions screening - casting a wider net to ensure no positive matches are missed.
Fuzzy matching algorithm
The algorithm used to determine this process is called Levenshtein Distance.
This algorithm has been tested extensively across different names and name variations in our supplier's database. To reduce false positives, the maximum edit distance change has been capped at one character. This allows for spelling errors/variations without returning large numbers of unnecessary false positives.
This article is based on accessing Thirdfort directly via our portal. If you access Thirdfort via a partner or reseller, functionality may differ.