Open danmaps opened 4 years ago
actually try this https://norvig.com/ngrams/count_1w.txt
instead of:
>>> sp.sentence("let's eat trail snacks")
["let's treat ail snacks", "let's treat ale snacks", "let's eat snail tracks", "let's eat snail trax"]
return:
["let's eat snail tracks", "let's treat ail snacks", "let's treat ale snacks", "let's eat snail trax"]
Idea: Instead of word popularity, try sorting by putting the "more perfect" (that work both on spelling and pronounciation) matches first.
potential resource: https://github.com/3b1b/videos/tree/master/_2022/wordle_data
another inspiring youtube video. Josh Wardle used the google ngram thing to rank words by popularity.
https://github.com/first20hours/google-10000-english