3 ms·
Filtering has worked pretty well with email for spam. I don't see anything fundamentally different about harassment that would make it less responsive to filter
by tzs 10y ago
Filtering has worked pretty well with email for spam. I don't see anything fundamentally different about harassment that would make it less responsive to filtering than spam, nor anything fundamentally different about an incoming social media stream than an email inbox.
Spam filtering works because the spam email has characteristics such as word frequencies that differ significantly from your non-spam email. E.g., none of my friends, acquaintances, co-workers, or companies that I correspond with ever have occasion to discuss my penis size, and so it was easy for my spam filters to learn that mail about my penis is spam.
I would expect that most people who get harassed on their social media feeds do not normally receive messages from friends or other wanted senders describing how they are going to rape them or kill their pets or children or things like that, so why can't a filter learn to block those?