I got the anki integration on Yomichan running and want to know what your main sources are for this method?
If you set up visual novels or games with a text hooker (something like textractor), it was super super easy to mine new cards with that without breaking the reading flow too much.
Also yomichan actually works with a lot of video subtitles too: both auto generated and actual subs on youtube for example. It also works with Substital, which I was using to add japanese subs to shows I was watching on other sites that donât have subs.
Then of course anything text based works great: any news articles, twitter, even discord conversations if you are running discord in a browser window instead of the desktop app version.
The mileage with automatically grabbing the surrounding sentences for the cards varies, since some of these will also grab a bunch of random junk (especially discord), but if you just want the word theyâre great
(My anki integration is broken currently and I havenât fixed it, but I did have it working for 1 click cards for a while)
I mostly look at my Android dictionaryâs history and type in with an IME on PC (inside Discourse, actually).
Bookwalker text-based books go first into Google Translate (for confirming the reading and for rough sentence segmentation and translation), then to clipboard history, then to the Android dictionary app.
Yeah, itâs double effort, but my Yomichan is more loaded than the Android app, and I donât want to add too many vocabularies to Anki. Also, I do some post-processing after adding, like adding Furigana (sometimes), images (sometimes), and context sentences (almost always).
Truthfully, apps like Akebi + AnkiDroid may simplify the process. Last time I see, Takoboto canât choose a proper form and a proper reading.
I donât post all vocabularies to the reading challenge. I deleted a lot.
Whoa! I didnât know about Akebi! Thanks! Now my phone is as useful as my PCs! @_@
Mostly news articles from Mainichi Shinbun and NHK. Nikkei is also a fine choice if youâre interested in more challenging topics.
Right now I am getting into Japanese dramas and more every day stuff. While I like those challenging topics in English, I am kind of at the Japanese High School level of interest lol
Hahaha, same! I do read about politics often enough, but my main interest is still high school smutty novels. Fortunately enough, there is a copious amount of these.
Iâm sure weâre probably pretty similar in that we are just kind of riding the wave to fluency lol. It is funny though because every year I seem to âgrow upâ a bit. Like 4 years ago I was struggling to watch Doraemon, and this year I am finally starting to lose interest in anime (or at least am running out of shows that genuinely interest me).
Unfortunately, for me thatâs been the case as of late . There are too many things to do and reading a more advanced book takes a lot of time still. For instance, Weathering with you is surprisingly not an ease read and itâs filled with words written in less common kanji. I have the book edition, though, which might arguably be harder.
Yeah, I guess in the end reading is where the fluency is at
Viki is underrated. Wish they had more series though, watched through most of them 2-3 times.
Take the time to understand which fields are sent to Anki, if you mess it up you may realize that youâre missing something later as you iterate over your card template
To be safe I just send every single field available in yomichan to Anki, then I just use the ones I care about in the card template. Better safe than sorry.
If you guys have Chrome-based browsers, I recommend using Language Reactor, you can have Yomichan make the flashcards off the stuff youâre watching too!
I made a yomichan package that shows the WK level for vocabulary and kanji:
I noticed that @Kumirei had already had the same idea in January of '22 but I donât know if they have kept it up to date with the various content shuffling since then, my package above has the data as of today.
Thatâs awesome! Gonna definitely add this to my cards!
Great idea, but I am fairly certain the ones by kumirei automatically update themselves using GitHub wizardry.
You seem to be correct, I couldnât find any version info in the package but I just noticed that çłăçłă was correctly reported as level 60 (which is a fairly recent change).
This topic was automatically closed 365 days after the last reply. New replies are no longer allowed.