This article is relevant for users who have terms consisting of 2 letters and a special character (e.g. R&D or I:E) in their termbases.

Term Recognition in Trados Studio does not recognize such terms. However, terms consisting of 3 letters and a special character are recognized.

This is a limitation of the term recognition. Such terms should be avoided.

Background on term recognition:

• It's based on trigrams and hence starts working fully once you have three letters or more.
• If those three letters are interrupted by any tokenized character (such as “&”, “-”, “:”), they are not seen as one entity anymore and will be split internally. This is by design to work properly with punctuation (e.g. for “:”) or recognizing hyphenated words (“-”).
• In most of these cases, alphanumeric token recognition will kick in and this is the best mechanism (not terminology) for handling such content elegantly.