so it could use a bayesian filter as well, although you'd have to be able to tune its sensitivity
I liked the idea of essentially running a spam filter backward. Instead of marking things š©, you š things, and it adjusts your filter.
Kind of like nostr:npub1utx00neqgqln72j22kej3ux7803c2k986henvvha4thuwfkper4s7r50e8 algo, but with the addition of examples, to the set of rules.
Discussion
Iām a little obsessed with long form articles, so I have ideas for favorites and semantic search working together