As Facebook and Google have become an essential part of our daily lives – a tool that cannot be missed, researchers around the world have been working on a way to identify fake news. Facebook has 20,000 people to monitor the content posted by its users. But there can never be enough to supervise all the phony stories.
Scientists at many institutions, including MIT, the University of Michigan and Clemson University, are trying to train computers to spot such fakes, using the AI. Marten Risius, an assistant professor of management at Clemson, had the same idea not long after the 2016 election. He and a colleague in Germany, Christian Janze, went to work on an automatic fake news detector.
Risius and Janze obtained a collection of more than 2,000 articles about the election posted on Facebook and compiled by the news site Buzzfeed.
Since fake news stories always based on the truth, they had their software look for common characteristic of stories. In the case of Risius and Janze, the found fake stories by spotting a large number of capitalized words and exclamation points. Clues can be found in the way other Facebook reader’s response to the news. It is likely that fake news tends to get more “love” or “laughter” icons but not many shares compare to the truth.
After its training, the software on another batch from the Buzzfeed compilation that it had not seen before, and Risius said it accurately identified the fakes 88 percent of the time.
Another effort to deal with fake news is Ramy Baly’s. He is a postdoctoral student at MIT and also develops software to identify trustworthy news sources on the Internet. Baly has spotted a host of subtle indicators to suggest a news site might not be on the level.
In December 2018, Boston Global Forum and Michael Dukakis Institute will also organize the Fourth Annual Global Cybersecurity Day, event title ‘AI solve Disinformation’, to explore the current state of cyber issues and the threat posed by disinformation and fake news, as well as effective defense mechanisms against these activities.