I am getting false positives for missing terms in QA verification with Chinese. This only seems to be happening when the expected term is a number:
For example, here the year 111 according to the ROC calendar should be translated as 2022. 2022 is in the target segment, but it is still flagging it as missing.
I don't think this is a "false positive" in terms of expecting this to work in the first place because Studio will not do this out of the box. Unless of course you have created your own custom QA rules to deal with the Minguo calendar? It would probably be better to have a custom plugin for this to avoid having to create a rule for each year since 1912. I also think it would be pretty trivial stuff for a developer to create one just for for this.
We can plan one in and if it's as simple as I think we'll try and prioritise it as a bit of training for a new developer. But no promises.
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
Paul Daniel Hug I’m not using a regex or custom QA rules. I just created a list of every minguo year from 1-150 and its expected translation 1912, 1913 etc. using ChatGPT and added it to my termbase. It works nicely in the term recognition, but unfortunately gives false positives when using the QA in Trados. For now, I’m getting around this using the XBench QA check which does not throw up these false positives.
Of course, custom QA rules for minguo years and other issues with Chinese numbers like months of the year would be really helpful!
Paul Daniel Hug I’m not using a regex or custom QA rules. I just created a list of every minguo year from 1-150 and its expected translation 1912, 1913 etc. using ChatGPT and added it to my termbase. It works nicely in the term recognition, but unfortunately gives false positives when using the QA in Trados. For now, I’m getting around this using the XBench QA check which does not throw up these false positives.
Of course, custom QA rules for minguo years and other issues with Chinese numbers like months of the year would be really helpful!
I just created a list of every minguo year from 1-150 and its expected translation 1912, 1913 etc. using ChatGPT and added it to my termbase.
So if you were to take this list and create the QA rules with it then you would be able to use your custom rules instead of the default, and then you would not have this problem.
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
Thanks, I will give it a try.
Translate
