I have been doing some quick tests with the MT Comparison plugin (https://appstore.sdl.com/language/app/mt-comparison/933/) and I have spotted some behaviors that I wanted to share with you, in case anything can be further improved in future releases:
- The MT Compare tool will generate a report with the MT providers selected during the project creation (e.g. ETS, BeGlobal, etc.). However, the report will not show which engine was selected in the Settings tab. Example:
Headings from the original report:
Headings manually modified:
- The tool does automatically give you the option to get the output from a BASELINE. However, it is not clear which engine that really is. For my initial tests with the baseline, I have used the language combination English>French and the output of the "BASELINE" column did not match the one I obtained with the beglobaltranslationprovider (which was pointing to the generic engine).
- The tool seems to be working well with plain text format, but rich-formatted files are throwing some issues. In particular, I have noticed that content between tags and placeholders does not get exported. EN>ES Example:
Original xliff file:
MT Compare Report: lines in red indicate that something has gone missing at that point. As can be seen in the last example, the removal of tags/content can also lead to glued terms.
- If you enable the same provider more than once (e.g. beglobaltranslationprovider) to point to several engines, the Report will only provide you with one output for that provider and you will not be able to tell to which engine it belongs to unless you process the file with each of the engines again. E.g.: in my project, I had the beglobaltranslationprovider 3 times (one to point to the generic engine, one to point to the travel vertical and one to point to an external engine). In the report though, I only have one column with the beglobaltranslationprovider heading.
Settings before generating the MT Comparison report:
MT Comparison report:
I hope this is useful. Please let me know if you need any further information. Thanks
We started working on the enhancements and problems you reported. Thank you for your feedback.
That's great, Emanuel Albu. Please do let me know if you need any further information. Thanks a lot for looking into it and for acting on it so quickly. Have a nice day!
Please read this: https://community.sdl.com/product-groups/translationproductivity/f/sdl_language_cloud/27171/lc-baseline-engine-produces-very-different-results-if-used-with-mt-compare-compared-to-using-pre-translate-batch-function---why
It's also about the MT Comparison App.