My name is Judyta, and I am a PhD student in linguistics. My students and I have been loving LLN almost since it launched, and I’ve recommended it as a great tool during a lot of my lectures.
Now I am using it as a research tool. I study subtitles, both human-made and machine-made. The question that popped up during my research is what kind of machine translation is used in LLN. To come up with any reliable conclusions, I would need to know whether machine translation in LLN is neural or statistical. I hope the developers can provide me with an answer, please!
For the sake of science, better outcomes and better translations (and my PhD sanity!)
If you have any additional questions about my research, don’t hesitate to ask.
Hi Judyta! I’m deeply glad to know there is someone like you ready to hear other people using this great tool to improve it more and more. Unfortunately I’m not a developer but my opinion could be a different point of view aimed to reach the same goal. In my opinion I can see most of choices of translated words from english to italian are right even if sometimes it’s difficult find out the right translation considering the immediateness during it has be translated a sudden unknown english word during the movie.
Thank you Luca for your comment!
However, it is actually key information to find out what kind of machine translation system they are using - in a scientific dissertation, it is not really enough to say “Oh, you see, this is correct, and this is not, but in general, it’s okay.” There are much more elements to it
So I truly hope to get the response from LLN developers