After translating an expression from English to French using NMT in Studio 2019, I made some changes to that translation. Later in the document, NMT translated an identical occurrence of the same English expression but without my changes. Is there a way that NMT can apply my changes from the first translation to the second one?
Thank you so much,
Karen
The trick is to use a Translation Memory. If you update your work into a Translation Memory then a similar segment will be handled by the Translation Memory in preference to the NMT.
Paul Filkin | RWS Group
________________________
Design your own training!
You've done the courses and still need to go a little further, or still not clear?
Tell us what you need in our Community Solutions Hub
Thank you so much for your answer, Paul!
The document I'm working on is indeed being saved in a new Translation Memory; how can I get it to handle a similar segment in this same document?
Please bear with me, I'm still a SDL studio newbie :)
You said:
how can I get it to handle a similar segment in this same document?
Initially you said this:
Later in the document, NMT translated an identical occurrence of the same English expression
Is the segment just similar, or is it exact?
If it's just similar to something you translated before you'll get NMT again, but if it's exactly the same you should get the match from your TM. Have you enabled your TM in your project settings?
Paul Filkin | RWS Group
________________________
Design your own training!
You've done the courses and still need to go a little further, or still not clear?
Tell us what you need in our Community Solutions Hub
Hi again Paul,
I now know what my problem is... The TM I created is actually empty. I have to do a bit more reading to learn how to transfer my confirmed segments to my TM. Thanks again for all your help and patience 
.
Karen
Perhaps it's something as simple as not confirming the translations (Ctrl+Enter) as you work? This will send them to the translation memory and move to the next segment as well as mark up each segment so you know it's been completed.
Or perhaps it's just not enabled in your project settings and checked to update:
Paul Filkin | RWS Group
________________________
Design your own training!
You've done the courses and still need to go a little further, or still not clear?
Tell us what you need in our Community Solutions Hub
Hi Paul,
This is what I've done. Not sure why I seem to have a subset for my TM.
Thanks for helping me!
Karen
Hi Paul,
This is what I've done. Not sure why I seem to have a subset for my TM.
Thanks for helping me!
Karen
Not sure why I seem to have a subset for my TM.
This is because you have created a "Project TM". This will be a subset of your Main TM containing only segments that are relevant to your translation based on your project settings for the match values. So if you set this to 100% matches only then it would only contain 100% matches from your Main TM, if you set it to 70% then it would contain 70% matches and above from your Main TM.
As a freelance translator it's normally better to work without a Project TM I think as the main usecase would be sharing a small TM with others.
The problem with using this is that your Main TM won't be updated until you run a batch task to update the Main TM or Finalize your project. It's also not the default so you must have specifically asked Trados Studio to do this when you created your project.
It's not a problem, it just means you need to know how this works and how to work with it otherwise you may think your TM is empty when in fact it might not be. Your project TM will be in the project folders if you open up the project in Windows Explorer. Then you can check it and see if it's empty.
You can also click on a segment in the Editor View and see if you get a TM Match for something you already translated. If you don't then can you share a screenshot of the screen as you see it when you do this check?
Paul Filkin | RWS Group
________________________
Design your own training!
You've done the courses and still need to go a little further, or still not clear?
Tell us what you need in our Community Solutions Hub
Hi Paul,
Yes, you're absolutely right! It's a project TM, I checked and it is not empty :). I made the mistake of creating my TM in my project instead of creating through the Translation Memories. I just have one last question: where do I set my matches to 70%?
Thank you!
Karen
In your Project Settings:
Paul Filkin | RWS Group
________________________
Design your own training!
You've done the courses and still need to go a little further, or still not clear?
Tell us what you need in our Community Solutions Hub
Thank you so much, Paul! Slowly I will get there!
Karen
But note... if you wanted to know where this setting was relevant to populating a Project TM then it would be here as you create your project:
The default is always 100%.
This is only relevant to the "Prepare" task. The one you probably want is this one:
This one will not prepare a Project TM, but it will use your pre-translate fuzzy value to populate your file when translating. So if you're using NMT as well you'd probably think about how much better NMT s than a low fuzzy and try to find the balance. So if, for your source material, you think a 90% fuzzy is always better than NMT then you would use 90% as the minimum match value... for example.
This article might be worth a read... pretty old now but maybe still relevant.
Paul Filkin | RWS Group
________________________
Design your own training!
You've done the courses and still need to go a little further, or still not clear?
Tell us what you need in our Community Solutions Hub
Thank you! Lots of reading and learning required
Translate
