feature request: naive baynes ham / spam classifier

[ag88]

it'd cool if this has a ham / spam classifier :)
e.g. if a user clicks a 'like' 👍 button classify it as ham a 'dislike' 👎 button classify it has spam
then have a filter the filter by ham classification, this can significantly reduce the reading list and still have 'all' for browsing