Kanji, every Japanese learner's worst nightmare!
There's really nothing more tedious than wrestling with 1,600+ characters, each with their own readings, pronunciations and meanings. Even after 10 years of study, it's still the wall I keep hitting. (And judging by the heap of flashcard apps out there, I'm definitely not alone...)
But there had to be some order to the madness, right? Some method to make this mass of characters less daunting. My own kanji journey's had its fair few bumps in the road, with, apps, and some painfully slow manga reading. The thing was, I know that I know kanji! I can recognise quite a lot of the characters, their meanings, and words they're in. I just can't read them themselves! Thus, I decided to step out of my comfort zone a bit, and utilise my background in linguistics with a less than perfect understanding of statistics, to see if I could group characters by both practical use and shared sounds.
The result was On'Yo! Using some statistical and AI magic, I managed to sort and group most of the jōyō kanji into a system that prioritises usability, daily importance, and proximity to other kanji readings. The idea is to use preexisting knowledge of other kanji as a structure to add new ones with the same readings, making it easier to connect and understand as you progress further down the line to more obscure characters!
It's still fairly rudimentary right now, and I've still got some ways to go before it's usable, but the stats and analysis are done! I'm hoping through this project, I can not only improve my Japanese proficiency (and coding skills), but maybe provide a tool to others who might benefit from this unique way of learning!