My most recent side project is meant to be a lighthearted thing with a dynamic subdomain where anyone can type [whatever-subdomain-they-want].is.obviously.best or [whatever-subdomain-they-want].are.obviously.best or [whatever-subdomain-they-want].is.not.obviously.best or [whatever-subdomain-they-want].are.not.obviously.best.

I have a list of political terms and people that route to an HTML page that says “[subdomain] has been flagged as political. The creator of this site intended this domain to be used to spread joy and merriment and feels that pushing political agendas undermines that intent.”

I have sentiment analysis in combination with a disallow list on is/are (positive, rather than is.not and are.not) routes that if the subdomain is flagged as negative by sentiment analysis or matches a term in the disallow list, it serves an HTML page that says “[subdomain] is/are NOT obviously best. What the hell is your problem?”

Sentiment analysis only goes so far and it’s hard for it to catch a lot of things (since it’s a small amount of input) and I’m not confident that I’ll think of all of the possible things that really shouldn’t resolve to is/are OBVIOUSLY best.

Is there anything you guys can think of that should be on the disallow list?

If it helps, the disallow list so far is https://raw.githubusercontent.com/A...