Don't you hate those Twitter accounts that spam you with friend requests or tweets that sound like someone is either really uptight or corporate-controlled? Well, those most likely are fake Twitter accounts that are controlled by bots, and now researchers found a way to identify them.
Network researchers at Indiana University have developed a tool, called BotOrNot, that analyzes more than 1,000 features from a user's friendship network structure, Twitter content and temporal information in real time.
The researchers come from the Bloomington School of Informatics and Computing, which is part of Indiana University (IU). It was funded by the U.S. Department of Defense "in order to counter technology based misinformation and deception campaigns," IU says.
It is interesting how the US govt. is focusing its efforts on finding misinformation and deception campaigns when it has been in the midsts of privacy scandals related to social media eavesdropping from the NSA. However, it makes sense for political campaigns that are becoming more and more social network driven for the govt. to want honest campaigns - and weed out potential shortcuts on campaign runners through algorithms or bots.
"The National Science Foundation and the U.S. military are funding the research after recognizing that increased information flow -- blogs, social networking sites, media-sharing technology -- along with an accelerated proliferation of mobile technology is changing the way communication and possibly misinformation campaigns are conducted," IU says.
Predictive analytics and big data are definitely changing the way organizations are learning about the social networks and online usage of their user bases. By analyzing data in real time and gathering analytics from it, predictive behaviors can be learnt, which in turn can help to find errors such as bots being present within networks.
"the 'secret sauce' is in the set of more than one thousand predictive features able to discriminate between human users and social bots, based on content and timing of their tweets, and the structure of their networks," said Alessandro Flammini, an associate professor of informatics and principal investigator on the project.
Keep in mind that although this tool is powerful it is still an estimate in identifying the bots. It identifies "most likely" Twitter accounts to be bots, but like any algorithm and computer program, it can make mistakes.
You may recall the good 'ol days of AOL chat, or even Yahoo chat, where you couldn't tell if you were talking to a real person or not. It seems that those days aren't totally gone just yet. Hopefully, this app will help weed out those accounts and have Twitter close them once they are verified to be bots.