I remember a time when you could safely assume that the person you were speaking to on Twitter was a human. Not, I’m afraid, any more. Twitter (and, more perhaps precisely, Twitter accounts) is now a weapon of war, used to spread misinformation and propaganda. And the most insidious thing is that it’s extremely difficult to distinguish what account is manned by a real human, and what one isn’t.
Botcheck.me is a machine-learning tool that makes the distinction for you, by observing how accounts act. If an account tweets constantly throughout the day, with messages just a few minutes apart, it’s likely a bot. If it tweets consist of sensational political messages and fake news, it probably isn’t kosher. Similarly, another major indicator is if the account obtained a large amount of followers in a relatively short amount of time — faster than you typically would if you used the service like a normal person.
The entire methodology and a further analysis are both available online, and are well worth reading if you want to understand how botcheck.me works on a deeper level.
In practice, to find out if an account is real, you just have to copy the username into the site. If an account exhibits the behavior of a political bot, Botcheck.me will let you know, while legitimate-looking accounts display thusly.
It’s a handy tool, not just for ensuring you don’t waste your time picking fights with algorithms, but also for seeing who is trying to influence your opinion. You can check it out here. There’s also a Chrome extension that’s makes the entire process even easier.
Get the TNW newsletter
Get the most important tech news in your inbox each week.