Im running SDL Trados Studio 2019 SR2 15.2.7.2849 and SDL Multiterm 2019 15.0.6.56482, both u`pdated to the latest versions
For example in Segment 146 term "Wasserfilter" is not displayed in the term recognition window:
But if I search "Wasserfilter" the termbase search window shows the entry:
How can I solve this problem?
Have you tried to create a TB in the language cloud and then conect it to your project through the Termbase option in Project settings?
I had the same problem and now term recognition works in Trados.
For the creation of the cloud TB you login to your RWS cloud and go to terminology, then you click on create new termbase, choose language pairs. After the creation you import the TB as excel file or another type.
Sadly, I tried this, and for me it didn't work. I still get the super annoying (and, as far as I can tell permanent, since the developers won't properly recognize it for what it is: a total mess) problem of terms that I know are in my term base not being recognized.
Every year or so, I open up Trados Studio and start a project, only to pretty much immediately discover that term recognition does not work reliably. Someone recently told me to try the cloud termbases, so I dutifully tried them a few days ago on an actual job. As usual, it didn't work, so I had two stop using Trados Studio mid-project and go back to memoQ, where terms are always recognized and highlighted if they are present in your termbase.
Term recognition has been broken in Trados Studio since time immemorial. I used to think it was caused by the crappy underlying technologies that the term bases are based on. However, since the cloud termbases also don't work, I wonder what's going on...? Sadly, I think the last person that will be able to tell you what's going on is the developers. I have never seen anyone from Trados admit that there is a very real problem and that they are going to fix it once and for all.
In short, if you rely heavily on terminology (as I do) while translating, don't waste your time on Trados. Use a CAT tool that works.
The number of hours of my life I have wasted on fighting with Trados, battling with the program, just to get the basics to work properly.
I'll create a new thread if and when I find a moment! Bt it really seems to be pointless, as I have been complaining about this problem for like 15 years. It truly boggles the mind that people put up with this.
I always use the latest, up-to-date versions of Trados Studio + MultiTerm, that's not the problem.
The steps to reproduce the problem are simple:
Create a project in the latest version of Trados Studio, create an online termbase (so no MultiTerm this time), and start translating. Add a few terms to the cloud termbase, and continue translating. A few segments down I find that (as usual), my Term Recognition window is empty. That's funny, I know I just added that specific term. Annoyed, I add it again. Still now showing. Lots of fun. Back to memoQ.
Over the years, I have followed everyone's well-meant advice (reorganise yr termbase, restart Studio, reset the Views, make sure the versions of Studio and MultiTerm match, and are updated, etc., etc., etc.), but none of it has ever worked consistently. It might seem to for 5 minutes, but not really. Indicating that all these "solutions" are just the user getting lucky. Term recognition in Studio has always been flaky and random. The only time I ever got rock-solid reliability out of a termbase in Studio was with that app by RyS (the so-called "RyS Termbase & Translation Assembler"; it's called something different now and is way overpriced).
all these "solutions" are just the user getting lucky
If that's true then I've been lucky for a very long time. I love using Trados and MultiTerm!
All the very best, have a good week!
Ali 
Looking forward to seeing the newest Studio 2024 to verify how they have evolved, if that's the case, the terminology department of their flagship application. MultiTerm, as I said it and proven already, was always sort of a dormant and underutilized feature of the Trados ecosystem.
Ha ha. I will eat my hat if term recognition works flawlessly in Studio 2024 (like in memoQ, Wordfast, CafeTran, Phrase, Déjà Vu, Across, OmegaT, MateCat, SmartCat, and Transit).
I will eat my hat if term recognition works flawlessly in Studio 2024
Shall I pass you the salt, pepper, ketchup? 
Apparently there are still problems. Terms with two letters and an ampersand are not recognized in the Term Recognition window. I had created an assistance ticket for it and received the following response:
The issue is indeed a product defect that is reproducible in the latest released Studio version as well (2024), logged under no. CRQ-20974. I will try to escalate the resolution of this product defect, but I cannot guarantee anything. Unfortunately, I cannot help further at the moment.
Apparently there are still problems. Terms with two letters and an ampersand are not recognized in the Term Recognition window. I had created an assistance ticket for it and received the following response:
The issue is indeed a product defect that is reproducible in the latest released Studio version as well (2024), logged under no. CRQ-20974. I will try to escalate the resolution of this product defect, but I cannot guarantee anything. Unfortunately, I cannot help further at the moment.
Terms with two letters and an ampersand are not recognized in the Term Recognition window
Hi Jonathan,
Are you saying terms with 2 letters plus an ampersand aren't recognised? Or terms with 2 letters, as well as terms with an ampersand?
I expect you know that you know that you need to have the following box ticked for recognition of 2-letter terms?
Regarding term recognition with an ampersand, are you saying that the terms in the termbase have an ampersand?
In Studio 2022, I found that false positives were returned for terms due to content that is around them or before or after, such as tags or complex punctuation...
I suspect that Studio 2024 will already be more accurate in that respect because the settings are now such that a Minimum match value of 100% really does mean 100%, therefore it flags up, for example, plurals as false positives. I'm finding that a Minimum match value of 95% stops that.
It's a work in process...
All the best,
Ali 
Alison Field Thanks for your reply. Yes, I know this setting and it is activated. In my case the source term 'S&O' (translation: l'équipe S & O) isn't shown in the Term Recognition Window, although it's in my termbase.
Another frustrating problem happens when a term that is written for example in 1 word (arbeidsongevallenverzekering) is written wrong in the source text (for example in two words: arbeidsongevallen verzekering). Suddenly, the term isn't recognised anymore (even when I lower the Minimum match value to 45%).
Hi Jonathan,
I wouldn't expect term verification to pick up a two-word source term where the termbase only has a single word. Also, I'd never go as low as 45% Minum match value (under Terminology Verifier > Verification Settings) because that would trigger a whole lot of unrequired false positives, because that figure applies to Source and Target terms.
If, however, you are referring to the Termbase Search Settings, i.e. the settings used when searching for source terms as you translate, then this match value can be lower and I find 70% usually finds everything I need except when the source term is split in to two words.
It would be good if you could create a new thread giving precise details of what you're flagging up as an issue, as I'm not sure here. Also I've just noticed how old this post is, over 4 years.
That way, someone from RWS (I'm just a Studio user, not an RWS employee) such as Paul can look at the issues you're describing and help you achieve greater accuracy in term recognition, etc.
All the best,
Ali
Hi Jonathan,
I wouldn't expect term verification to pick up a two-word source term where the termbase only has a single word. Also, I'd never go as low as 45% Minum match value (under Terminology Verifier > Verification Settings) because that would trigger a whole lot of unrequired false positives, because that figure applies to Source and Target terms.
If, however, you are referring to the Termbase Search Settings, i.e. the settings used when searching for source terms as you translate, then this match value can be lower and I find 70% usually finds everything I need except when the source term is split in to two words.
It would be good if you could create a new thread giving precise details of what you're flagging up as an issue, as I'm not sure here. Also I've just noticed how old this post is, over 4 years.
That way, someone from RWS (I'm just a Studio user, not an RWS employee) such as Paul can look at the issues you're describing and help you achieve greater accuracy in term recognition, etc.
All the best,
Ali
Translate
