Glossary

Fuzzy matching

Fuzzy matching is a translation memory (TM) feature that identifies segments of text that are similar – but not identical – to previously translated content. It helps translators reuse existing material efficiently by showing partial matches between a new segment and stored translation units (TUs).

Description

When a computer-assisted translation (CAT) tool performs a fuzzy match, it compares new text against entries in the translation memory and assigns a similarity score, usually between 50% and 99%. This score reflects how closely the new content resembles a previously translated segment.

For example, a sentence differing by one or two words may produce a 90% match. Translators can then edit the suggested translation instead of starting from scratch, saving time while maintaining consistency in tone and terminology. In RWS Trados, fuzzy matching is central to improving translation productivity. It works in tandem with concordance search, TM leverage (translation memory leverage) and AI-assisted translation features to accelerate project turnaround while upholding linguistic quality and client-specific terminology.

Related terms

CAT tools (Computer-assisted translation) Concordance search Linguistic quality Post-Editing TM leverage (translation memory leverage) Terminology management Trados Team Translation memory (TM) Translation unit

Transform how the world understands you.

Discover the technology and expertise that powers the world’s most trusted brands.
See what's possible