When it comes to the war on misinformation, artificial intelligence may be as effective, if not better, at spotting fake news stories than humans, new research has found.
According to a study, researchers at the University of Michigan have developed an algorithm-based system able to detect propagated news on internet-based media platforms like Google News and Facebook by analyzing particular telltale linguistic cues.
It found fake stories up to 76 percent of the time, compared to a human counterpart, which found a success rate of 70 percent. The algorithm-based system may also be efficient enough to spot fake news stories before they are cross-referenced with other stories.
The algorithm, in which researchers concluded may at times work faster than humans at linguistic analysis, was programmed to examine several attributes in stories including word choices, grammar usage, and punctuation.
For the study, researchers recruited participants to rewrite 240 news stories. The style of each story was to be mimicked but rewritten into a fake news article. The algorithm then analyzed both versions and was able to successfully spot 76 percent of the rewritten stories.
Rada Mihalcea, a computer science and engineering professor at U-M, said such automated tools can be crucial for news providers that struggle to keep tabs on misinformation, as many still rely on human editors.
“You can imagine any number of applications for this on the front or back end of a news or social media site. It could provide users with an estimate of the trustworthiness of individual stories or a whole news site. Or it could be a first line of defense on the back end of a news site, flagging suspicious stories for further review. A 76 percent success rate leaves a fairly large margin of error, but it can still provide valuable insight when it’s used alongside humans.”
The paper, titled “Automatic detection of Fake News,” was presented August 24th at the 27th International Conference on Computational Linguistics.