So i contacted the support for like 2 weeks ago with problems with words that says 1 thing when you play the audio and when you press the word to translate it, it says a whole other word.
I stopped doing the famous Mini stories because the translations to swedish on the mini stories was inaccurate , if it just was the grammar i could have dealt with it but the translation to what was being said in japanese and the translation to swedish was not even close. I even tried to go into edit mode to fix it. But because its lingq made i cantā¦
I also notice that many kanji characters dont have the full Hiragana as helpful words to read the kanji and or give the wrong hiragana for the word. like the kanji for āIchiā says āHitoā and āhitoā has ājinā and āWatashiā says āWatakushiā etc. etcā¦ there are so many things that are out of place.
I have also noticed when i import lessons that the program does not accurately seperate the words as it should, i have to go into the lessons and redo all the mistakes it has madeā¦ Even when i have seperated the words myself for short stories the program still seems to want to break the words down even more to make it unreadbleā¦
Is this just a problem the Swedish community here on lingq or are more people here having the same problem? kinda feels sad that i even payed for the premium when there are so many translations, sound, kanji readings issues.
The problems are known. The transliterations are auto-generated and therefore error-prone. If this is important to you, I suggest you copy the correct one (from a dictionary) into the meaning field, that way you are in control and you donāt have to rely on an algorithm.
The word splitting issue affects Chinese and Japanese here on LingQ because these languages donāt use spaces to separate words. As a result LingQ has to employ what is called a word splitter or word tokenizer to segment a string of characters into individual words. The results are far from perfect. LingQ may update the software at one point but since weāre dealing with a computer program this unlikely ever to be perfect. Correcting these errors can take a lot of time that isnāt productive study time. Iāve typically created āphrasesā in these cases, but that feature has been pretty buggy recently. Iāve written more on the topic of word splitting here: Chinese Word Recognition (Wrong Characters Are Combined) ...
The Mini stories are indeed locked to editing. When they get first added, the translation fields are empty. So, when the first user comes along and clicks translate in sentence mode, LingQ will ask Google Translate and save the result with the story. Any user requesting a translation at any later point will be presented with the GT translation. It doesnāt surprise me that these arenāt great. Ideally LingQ would add proper, human made, translations for all of their content into all possible languages, but Iām not sure theyāre willing or able to do that.
Anyways, the LingQ experience in zh/ja isnāt as smooth as with Western languages, I donāt expect that to change. But Iāve found LingQ to be a valuable tool, in spite of all the bugs.
Thanks for the reply!
Iāll just have to bite the sour apple and go with it. But yes i do agree the platform is valuable to get lots of exposure to the targeted language.