Twitter to use an algorithm to weed out abusive accounts
Twitter is going to rely on an algorithm to weed out accounts engaging in “abusive behaviour”. Abusive behaviour in the Twitter lingo is defined as either “repeatedly Tweeting without solicitation at non-followers” or violation of its terms and conditions. Social media analysts had expected that the change would be confined to swear word policing, but this one is much complex and better refined. It keeps into account the relationship between two accounts and hence, the people managing it.
If a person is engaging in such abusive behaviour he or she may be penalized by making his tweets visible to only his followers, forcibly making him verify his phone and mail address or he might also face suspension from Twitter for 12 hours. As per a statement by the company’s engineering vice president Ed Ho, Twitter is in favour of freedom of expression. However, in cases of repeated violation of the terms of conduct, it will consider taking further action.
The new algorithm is expected to face some hiccups and Ho acknowledged the same. He wrote, ““Since these tools are new we will sometimes make mistakes, but know that we are actively working to improve and iterate on them every day.” As of now, users will not have an option to appeal against the penalties.
A filter option will also be introduced wherein users will be able to filter notifications from accounts which do not have a profile photo or the ones which do not have a verified e-mail address or phone number. This step has been a result of observations that abuse is mostly, a by-product of accounts of this kind. Users will also be able to mute accounts, keywords or conversations for a specific period of time. These features resemble the functionality appended to verified accounts belonging to the social media platform.
If you decide to report a tweet, you will receive notifications regarding progress on the same. Previously, the company didn’t communicate to the users on whether it took any action against harassment complaints. This resulted in users being clueless on whether they have put through their concerns successfully.