In this repository, I compile a Language Training Toolkit to scrape resources of information and programatically make training material.
- Polish
- given annotated text, strip out words for inflection practice
- having a place to get usage data will help me crawl better, and also tell me which words are better to prioritize in terms of adding it to vocabulary and deciding how much to train it with flash cards
- add more languages
- add practice for characters
- I think some terms have partially completed entries, i.e. lata - am I handling this? If I find lata, do I end up finding rok?
- How can I tailor the inflection studying to the user? I.e. can I identify what forms the user needs to study, and then give the user more of those forms?