The lede: “Twitter Inc. on Wednesday launched a wider effort to use algorithms to identify accounts as potentially engaging in abusive behavior, a departure from its practice of relying on users to report accounts that should be reviewed for possible violation of its rules.”
From Reuters via globeandmail.com (“Twitter turns to algorithms to clamp down on abusive content“):
Twitter and rivals like Facebook have long relied on users’ reporting potential abuse for review, sometimes to the chagrin of groups that accused them of doing too little to thwart hateful speech or harassment. Twitter, which already uses technology to try and limit some communications, will still review user reports on potential abuse.
Twitter said it will limit the functionality of accounts flagged by its technology as abusive for an unspecified amount of time, a restriction that could include allowing only followers to see that user’s tweets. Currently, accounts are deleted or suspended when marked as abusive.
“We aim to only act on accounts when we’re confident, based on our algorithms, that their behavior is abusive,” Ed Ho, vice president of engineering, wrote in a blog post. “Since these tools are new we will sometimes make mistakes, but know that we are actively working to improve and iterate on them everyday.”
Twitter is also introducing new filtering options for notifications to allow users to limit what they see from certain types of accounts, such as those that lack a profile photo, and said it would alert users when it received abuse reports and inform them if further action against certain accounts takes places.