Google’s Official Profanity API(labs.phurix.net)
labs.phurix.net
Google’s Official Profanity API
http://labs.phurix.net/posts/googles-official-profanity-api
11 comments
So when the title says official, they actually mean "internal api, dug out of javascript source on a google affiliated site"?
What the {"response": "true"} is this {"response": "true"} thing? A {"response": "true"} cache page? Are you {"response": "true"} {"response": "true"} me?! Do you really need a {"response": "true"} JSON API to do the work of a stupid {"response": "true"} array lookup? {"response": "true"}!!!
Why oh why does profanity ever need to be filtered out?
> you could try using the Google Profanity API to avoid displaying ads on pages that contain profanities
I assume that must be a requirement of some ad networks
I assume that must be a requirement of some ad networks
Possibly because there's enough of it that it would affect statistical algorithms involving language processing (e.g. google translate)
I once tried to create a list of basic patterns...always thought it'd be more useful than it turned out to be:
https://github.com/mogade/badwords/blob/master/en.txt
https://github.com/mogade/badwords/blob/master/en.txt
CPAN, as always, to the rescue:
http://search.cpan.org/~abigail/Regexp-Common-2011041701/lib...
http://search.cpan.org/~abigail/Regexp-Common-2011041701/lib...
It's repetitive - when you're checking e.g. for (ph|f)u(c|k|ck|q), there's no need to check cyb[e3]r(ph|f)u(c|k|ck|q) as well...
> fixed some broken patterns, added check for dick
My stepfather would be most displeased if he couldn't use his name.
My stepfather would be most displeased if he couldn't use his name.
5hit, that's a fuhked up list.
basically:
http://www.wdyl.com/profanity?q=xxxNice, and it passes a couple of examples of the Scunthorpe Problem: http://en.wikipedia.org/wiki/Scunthorpe_problem
My first thought was "Oh, I wonder if that means the spellcheck on my smartphone will stop substituting 'duck' when I meant to type something more profane." But after reading the article, I guess not.
A maybe condition is necessary.
see: http://www.wdyl.com/profanity?q=balls
see: http://www.wdyl.com/profanity?q=balls
Yeah. Even with added SFW context:
> http://www.wdyl.com/profanity?q=tennis%20balls
> {"response": "true"}
> http://www.wdyl.com/profanity?q=tennis%20balls
> {"response": "true"}
It's publicly accessible, but I wouldn't call it official.
server dead, not in google cache
This Google cache link works for me: http://webcache.googleusercontent.com/search?q=cache:labs.ph...
here is copy of the post: http://pastebin.com/8C3EXMZ7