You are currently working on UAT 

How can we penalize the segments already pretranslated in the sdlxliff files that are not in memory?

I would like to know if we can penalize the segments already pretranslated in the sdlxliff files by our customer that are not in memory, so that we can give priority to our TMs.

These are the steps we are following: 

  1. We create a new project with all the SDLXLIFF files adding our own TMs to the new project.
  2. There are segments in these SDLXLIFF files which are already pretranslated by our customer. So, we apply a filter to search all the segments with a 100 % match and change their status to “draft”.
  3. After that we select the option “Pretranslate” to pretranslate the whole project using our TMs. And here is the problem: the segments pretranslated by our customer are selected rather than the translated segments of our own TMs.
Parents
  • Hello, Greta Magyar.

    We select the option "Always overwrite existing translation" including also the option "Overwrite PerfectMatch translations" and this does not work with all the segments as I told you before. We don´t know how we can solve it.

    The answers to your questions:

    - We have to keep those translations only when we decide they are correct. Many of the segments pretranslated by our customer need to be modified.

    - They always use pretranslation and won´t stop doing so.

    I hope someone can help us.

    Thanks anyway for your answer.

    Best regards.

    Melissa

  • Hello Melissa,

    Apologies for only getting back to you now.

    I have another suggestion, which might be useful:

    1. you could upload the content of the pre-translated files as received from the customer into a separate TM,
    2. then delete the translations in the files, empty all target segments,
    3. and then pre-translate the files again with your TM first in the list (with your translations) and the TM created with the customer's translations as the second in the list, with a decent penalty (the percentage is up to you, I would first test it with maybe 3%).

    This would ensure that your own content is prioritized over that of the customer. What do you think?

    Best wishes,
    Greta

Reply
  • Hello Melissa,

    Apologies for only getting back to you now.

    I have another suggestion, which might be useful:

    1. you could upload the content of the pre-translated files as received from the customer into a separate TM,
    2. then delete the translations in the files, empty all target segments,
    3. and then pre-translate the files again with your TM first in the list (with your translations) and the TM created with the customer's translations as the second in the list, with a decent penalty (the percentage is up to you, I would first test it with maybe 3%).

    This would ensure that your own content is prioritized over that of the customer. What do you think?

    Best wishes,
    Greta

Children