The Twitter algorithms that rank tweets will be checking negative tweets that include below-mentioned factors-
- Tweets that are sent by accounts with unconfirmed email addresses.
- The multiple accounts made by the user sending the tweets.
- Accounts that send tweets, again and again, mentioning other accounts which aren’t following back.
- IP addresses.
- Whether a specific tweet encouraged users to block the account that has sent the tweets.
When not considering the main content of tweets, Twitter can bring these changes on global level because language will not be a restriction.
Instead of completely removing negative tweets, Twitter will be filtering them by keeping them behind a “see more replies” button.
Therefore, in a test done, the results show that filtering negative tweets made a 4% decline in abuse reports from search results and 8 % reduction in abuse reports from chats.