Show HN: AI‑curated actual profanity list

github.com

1 pointhypernewbie2 comments

I made VBW, an AI-filtered multilingual profanity list that focuses on actual abusive language, not stuff like “farted,” “willy,” or “小姐.”. It’s intended for light content moderation, e.g., filtering usernames.

Check it out here: https://github.com/hypernewbie/vbw

It's not complete, some obvious words are still missing, but I hope someone finds it useful.

Open articleView on HN

> sv.txt

I wouldn't consider many of those terms as profanity, and especially not "very bad".

Yeah if you're looking at sv.txt, completely agree. That's the source depot I'm aggregating from before filtering. The actual filtered output is in vbw.csv.

A lot of the zh.txt and en.txt have the exact same problem. That's exactly why I made this project, to collect thesr and filter out.

Maybe I can push the filtering more.

Reply on news.ycombinator.com