Northonlifelock (formerly Symantec) last week released a beta version of a free browser extension that allows Twitter users to identify bots on this social platform. The tool is named BotSight and can currently be installed in Google Chrome, Brave, and Firefox browsers. The developers also plan to create a corresponding mobile app and a version of the addon that will work with Microsoft Edge.
According to the NortonLifeLock team, BotSight uses a machine learning model to detect bots on Twitter, and the tool demonstrates a high percentage of accurate detection of such accounts. As a result, users will be able to distinguish bot posts in the feed, search, and trends section. To identify a robot account, BotSight analyzes more than 20 features of the account: name, description, and speed of subscriber recruitment.
After installing the extension, users will see percentages and icons in three colors next to their accounts: green, yellow, and red. Using these indicators, you can see how likely it is that you are communicating with a bot (or reading its tweets).
To date, the developers have analyzed more than 100,000 accounts and concluded that about 5% of tweets are published by bots.