Learning Jouyou Kanji, a little Analytics, backend updates

A little bit of collapsible analytics and A Jouyou kanji deck build in

Also, mobile screens

I have spent a little bit of time working on the kanji Study feature trying to get it to the point where I can find it useful. For me that means at least a built-in study deck and some analytic. And that is what we have here…

The data for the bars in the photo is mock data at this time, but the API endpoints exist and work, I just don’t have a deck with pretty data ATM.

I build a collapsible analytics page that shows with colored bars and individual color-coded characters, which show where each flash card is due in each deck.

The cool part is that if you are studying kanji on your own in your personal study deck this progress will automatically be synced into your Jouyou kanji decks and vise-versa, so you don’t have to worry about studying something twice when you are studying your personal kanji decks you make from the dictionary.

The jouyou kanji decks are divided by level but really have no order inside the level yet. This is really something I hope to address, but it seems that after grade one, the order might not matter that much, but this is still research I will have to do.

I also have somewhat wasted a few days trying to speed up the dictionary lookup, as most searches take around 3s. I wrote some queries that save off the pure result Json for each dictionary entry at the gloss level, and found I could get almost instant results and a similar search order. At first, this was at the wrong aggregation level unfortunately and already took 40mins to generation at around 12500 records. At the correct aggregation level, I would produce around 420,000 records, and after trying to run the correct statement for 4hr, it failed as my dev machine hibernated… so I will shelf this for a while too.

I think next I want to focus in on some data errors I have in my JLPT word list, and start on the UI for the output of the media processing pipelines, so I can start viewing the data for the anime words in the app and start linking them to the dictionary.

Progress made is good.

Oh yeah and and it all should be working on mobile screens as well now too