(Sound effect: sound of robotics and typing on keyboard) Bot--or not?
I'm Bob Karson with the discovery files--new advances in science and engineering from the National Science Foundation.
Is a twitter account being operated by a real person or by automated software known as a "social bot?" Complex networks researchers at Indiana University have come up with a tool that separates the friends from the bots: A system called 'bot or not.'
The system's 'secret sauce' is a set of more than a thousand predictive features able to discriminate between human users and social bots. It analyzes certain patterns, language cues, timing of tweets even what's called "sentiment data" from content spreading through social media. The result: One powerful spot-a-bot system. In tests it could uncover a likely bot correctly 95 percent of the time. By the way, the social bots were provided by Texas A&M's infolab.
The work of the Indiana team is part of a larger program to support human efforts to counter misinformation with truthful information.
So how many social bots are out there? Nobody knows. But the researchers have come across "nasty" bots used they say, "to mislead, exploit and manipulate discourse with rumors, spam, malware, misinformation and slander."
I feel like we're in a movie trailer for..."attack of the twitter-bots!" (Sound effect: woman screams in panic)
"The discovery files" covers projects funded by the government's National Science Foundation. Federally sponsored research--brought to you, by you! Learn more at nsf.gov or on our podcast.