A new software system, BotSlayer, was designed to help people identify whether or not something found on Twitter was sent out by a bot. This software was created at IU by the Observatory on Social Media, a research center focused on the spread of misinformation on social media.
“A lot of people access information and news through social media,” IU informatics professor Fillipo Menczer said. “So, if you can manipulate what they see online, you can also, to some degree, control their opinions, their beliefs, their thoughts, and perhaps how they vote.”
BotSlayer is based on other tools OSoMe has had in the past, like Hoaxy, which lets people track fake information, and Botometer, which shows how likely an account is to be a bot.
These past tools have helped in picking out misinformation in elections. In the 2016 elections, Menczer said OSoMe was able to identify how articles were pushed out on social media and how likely human users were to share them.
Informatics PhD student Pikmai Hui said after going through multiple processes of analysis with these previous softwares, OSoMe realized the processes were similar. Even when the focus of research changed, the process from data collection to data storage to analyzing the data in a certain way remained similar. This prompted the thought that maybe they could automate the process so it is more beneficial for users, like journalists or researchers.
In the 2018 midterm elections, other people using IU’s software reported fake accounts on platforms like Twitter and Facebook, and got those accounts taken down.
Since OSoMe has released BotSlayer, over 120 organizations, including the Associated Press and the New York Times, have shown interest in the software.
Another student who worked on the software is a computer science PhD student, Christopher Torres.
“So now that we have provided them a tool to do this on their own that doesn’t require too much technical background knowledge, it’s really exciting,” Torres said.
Menczer said he hopes BotSlayer can help in the 2020 elections. In order to use this service, organizations must sign an agreement that says they will also follow Twitter’s agreements.
Torres reminds people not to believe everything they see online.
Hui said the transmission of information through social media is more the cause of the spread of misinformation than the actual source of the information.
“It reminds me how important it is to help people gain media literacy so they learn how to protect themselves from this misinformation when they see it with their bare eyes,” Hui said.