Terminology Verification giving too many false positives

Hello! I would like to know if there is any way to force the terminology verifier in Studio to not flag certain terms as errors if the stem of the term is consistent.

I am referring to terms where the stem of the term is the same, and only the ending is different. In Greek we have many word declensions, and this makes the verifier go crazy...

For example, for the term "win", I could have the basic translation "κερδίζω", and then variations such as "κερδίζουν", "κερδίζουμε", "κέρδος" and so on (the stem is the same, irrelevant to where the accent is placed, btw).

If I run a QA with the term "win" as "κερδίζω" in Multiterm, then somewhat all the words 'not' "κερδίζω" are flagged as errors.

I know that, for example, memoQ has the ability to create a glossary list with special characters inside the terms, indicating what the "stem" of the term is and what the software can safely ignore. For example, for "κερδίζω", I would put this term in the glossary like this: κερδ| or κερδ* (I don't remember the exact symbols, sorry).

This way the qa check would ignore all the endings after "κερδ" and not flag words like "κερδών" as errors.

Do we have that option or anything similar in Studio? Is there a way to make a special checklist that would "ignore" the accent wherever it is and just consider the letters? For example, consider sýmbol and symból the same.


Thanks in advance!

Sample terms (Ι made the stem of the term bold, for your reference)

symbol
σύμβολο

In some cases, the accent will drop, so the resulting stem will be like this: συμβόλ

reel
κύλινδρος

In some cases, the accent will drop, so the resulting stem will be like this: κυλίνδρ

Sample source and target text

For 3 Scatter trigger, the stack of symbols is selected from:
Για ενεργοποίηση 3 Scatter, επιλέγεται η στοίβα συμβόλων μεταξύ των:

Symbol upgrade progression is displayed below the reels
Η σειρά των αναβαθμίσεων συμβόλων εμφανίζεται κάτω από τους κυλίνδρους

Blue symbol upgrades apply to all instances of the respective symbols on all reels
Οι αναβαθμίσεις μπλε συμβόλου ισχύουν για όλες τις περιπτώσεις των αντίστοιχων συμβόλων σε όλους τους κυλίνδρους

Symbol Upgrade!
Αναβάθμιση Συμβόλου!

Symbol upgrade progression is displayed below the reels
Η σειρά των αναβαθμίσεων συμβόλων εμφανίζεται κάτω από τους κυλίνδρους

 

Parents
  • Former Member
    0 Former Member

    So, I'm thinking like this.
    Basically, my purpose is to make the whole procedure a lot easier.

    The procedure goes like this.
    1. currently you have a TermBase
    2. but, a file has "a little bit" of different terms from the TermBase
    3. so, you got false positive alarms
    4. if you expand your current TermBase according to your current file (that means, add the term pairs appropriately, nothing mystique whatsoever)
    5. you should be happy with it
    6. Real problem here is, above step #4 is very painful and waste lots of time and energy (News says everyday, globally all crazy hot weather..)

    Right?

     

    the upper one is Termbase before expand. What you are in currently.
    the lower one is Termbase after expanded (automatically, I hope). What you needed.

    Screenshot of Trados Studio showing a TermBase with terms highlighted in green indicating no issues before expansion.

    emoji


    Generated Image Alt-Text
    [edited by: Trados AI at 1:54 PM (GMT 0) on 28 Feb 2024]
  • Former Member
    0 Former Member in reply to Former Member

    a little bit more elaborated my workaround.

    If you make this MS Excel file using TermBase and docuemnt easily, your agony should be disappeared let's say 97% ?

    Because you just clicked once and the way to go is another several clicks only to make newer (expanded) TermBase.
    -I believe all the items more than 75% or something will give you a finer satisfaction.

    In the screenshot the Yellow part is parsed from the document automatically.
    And I believe you are familiar with the rest steps (make new TermBase from Excel file).

    Screenshot of an Excel file with a table showing source and target language terms, highlighted 'Friends' column indicating term match percentages, and a 'percentage' column with values.  

    I call this project as "Find Friends" (70% or higher friendship)

    regards

    emoji


    Generated Image Alt-Text
    [edited by: Trados AI at 1:54 PM (GMT 0) on 28 Feb 2024]
Reply
  • Former Member
    0 Former Member in reply to Former Member

    a little bit more elaborated my workaround.

    If you make this MS Excel file using TermBase and docuemnt easily, your agony should be disappeared let's say 97% ?

    Because you just clicked once and the way to go is another several clicks only to make newer (expanded) TermBase.
    -I believe all the items more than 75% or something will give you a finer satisfaction.

    In the screenshot the Yellow part is parsed from the document automatically.
    And I believe you are familiar with the rest steps (make new TermBase from Excel file).

    Screenshot of an Excel file with a table showing source and target language terms, highlighted 'Friends' column indicating term match percentages, and a 'percentage' column with values.  

    I call this project as "Find Friends" (70% or higher friendship)

    regards

    emoji


    Generated Image Alt-Text
    [edited by: Trados AI at 1:54 PM (GMT 0) on 28 Feb 2024]
Children