Hello all, I’m a beginner Japanese learner. My knowledge mostly comes from studying Genki 1 Textbook halfway, I also used Anki from time to time for Vocabulary/Kanji.
I want to make an iOS Spaced Repetition System app that builds a deck from any piece of text so that I can study from the content that interests me.
Here is how I want to do it.
Highlight text anywhere on your phone and click a process button
Now you have flashcards made from that source
the app tracks what you already know so it won’t make a card for it twice, can also mark already known words
Oh and a light scheduling tool so you are notified daily to do it (I myself don’t like notifications) but want to know if this is helpful to anyone.
I did not build anything yet and wanted to kindly ask for a reality check from the community here. Is making a study deck from any piece of text (on your phone) something that would help you in your studies, or is this just a gimmick?
1 Is “just” OCR from an image. Plenty of tools exist that gives you raw japanese text from an image. You can use tooling like GitHub - matt-m-o/YomiNinja: Open-source OCR and dictionary tool. · GitHub / mokuro etc… or if you’re tech savy (and want to script a workflow) you can run an OCR model locally (I’m doing so with docker and it works pretty well even if there’s some postprocessing to do after, using PaddleOCR-VL, specialised for vertical text). There are many tools existing… your use case is probably covered.
2 to 4 are already covered by services like https://jpdb.io/ | https://jiten.moe/ | https://hanabira.org/ … (according to what i understand from your description) you just input in custom decks of cards the raw text from 1.
So in general my advice would be to get to the “learn japanese” part and not reinvent those tools.
