Hey Lingq community,
Just wanted to express my thanks for the Lingq team for being one of the only few to create an app where Vietnamese is more learnable now in comparison to before. Definitely a huge upgrade. The auto pairing with words is a huge game changer and saves lots of time
Yes, let me add my thanks here too. A major step forwards for Vietnamese.
If anything, it goes a little too far in grouping words. For example it shows người thân thiện (friendly people) as as a single group, whereas I would be happier breaking người (person/people - a generic noun) and thân thiện (friendly - adjective) apart. But maybe this isn’t easily controllable and overall I am much happier with the new approach and looking forwards to resuming my Vietnamese learning with Lingq. For me, it is absolutely the difference between Lingq being unusable and readily usable for learning Vietnamese.
I’d be genuinely curious to hear how much you use LingQ for Vietnamese and what your level is, or to hear from other Vietnamese users. Because I’ve used it extensively as my main tool for Vietnamese and this new update has been quite frustrating for me.
I understand that the lack of word grouping was a major issue for beginners, so I can see why this feature was introduced. However, once you move past the beginner stage, it can actually be useful to click on the individual morphemes, and it becomes easy to recognize bi-syllabic words.
As the previous comment mentioned, this feature currently groups far more than what I would consider actual words. For example, it groups items like “những” or “này” together with the nouns, or “phản ứng hóa học” = “chemical reaction”, but this really should be two words. That makes it impossible to look them up properly in a dictionary and also unnecessarily multiplies the number of LingQs you need to create.
It also sometimes identifies nonsensical “words” when there are spelling errors (which are very common in YouTube subtitles), again making it difficult to check the correct dictionary definition. Even worse, in some cases it seems to go completely off the rails and group entire sentences together (see the examples below).
Used Vietnamese the moment it came out 2 years ago. Was probably one of the first few users to test it as well. The grouping would be around 80-90% accurate since it’s trying to involve all possible words possible in the past, future, and present tense. Started using the app at a A2-B1 level and reached around a B2 now with the assistance of lingq. But technically nowadays, language reactor is still king in terms of software for vietnamese while listening but lingq makes it superior in terms of learning vocabulary to increase the potential for learning the language at a faster and efficient rate. Both combined is king.
If you check out my personal stats, it will show the amount of work required to reach a decent level. It’s still a fairly alienated and difficult language.
Maybe it could be made turn off-able, if it isn’t at the moment? So beginners can use it but once they reached a certain level, where it becomes more of a hinderance as a help as you described it, can turn it off. 
Yes, exactly, we need to be able to turn this feature off.
LingQ team, can you explain why it wouldn’t be possible to provide an option to disable it?
I agree that the changes to group words together in Vietnamese makes things much better now than when Vietnamese was first added, however it seems to be really over zealous in how it groups some words now.
An extreme example I just came across: “em cũng luôn nói với học sinh” (I also always speak with students) as a single word?? I appreciate sometimes it’s difficult to make an exact judgement especially with whatever automated system is being used, but this one seems very clear-cut. It should be at least 6 separate words. This makes the known-word percentage very difficult to trust as you’re constantly adding “new words” that are actually whole phrases.
I hope more work can be put into improving this in the future. Thank you!
