Oh this is really awesome. I'm so curious how they calculate the scores. Proud of my 3% score- - but even then, my "worst" offenses were not actually haterisms.
Idea: Allow users to report non-offenses so that the algorithm can improve over time based on these reports. Sort of like how users who report spam help spam-fighting tools get better.
It would also be kinda funny / kinda cool to have an Internet's Most Wanted list of top trolls of all time.
@KevinMcAlear Maybe crowdsource it, have a landing page for people who want to "Report a Troll" and then rank the top ones of those :) Dunno just an idea. Good luck anyway, and holler if you want more feedback @narekk on twitter.
If anyone would like to know more about how it is built I wrote up the basics of how the algorithm works. and my findings in general. I'll keep working on it and if anyone has any suggestions let me know! :)
http://kevinmcalear.com/thoughts...
Also, if you want to help out with it here is the repo:
https://github.com/kevinmcalear/...
Surge
Solvemigo - ChatGPT for Telegram
1Degree
Channel
1Degree
Channel
Channel
Healthfundit
Channel
Channel