Researchers at Carnegie Mellon University have developed an algorithm that detects and flags stories that are clearly fake.
The findings, published in the journal Science Advances, may lead to better social media screening for people with autism spectrum disorder, which includes autism, Asperger’s, and pervasive developmental disorder.
“The most common way people fake news is through social media,” said lead author Andrew Miller, a Ph.
D. student in computer science at the university.
“This study provides a way to identify the fake news and prevent it from spreading through the social networks.”
Miller and his co-authors analyzed the social media activity of more than 5,000 people from the autism spectrum, including people who had a diagnosis of autism spectrum disorders, autism spectrum symptoms, and Aspergers syndrome.
The team identified and flagged 3,521 of those articles that were likely to be fake.
That included articles that purported to be news from an organization with links to the company’s own website, but were instead based on fake content that was created by an individual or a group of individuals.
“We found that the fake content was more likely to originate from the people who created it,” Miller said.
“It’s not the content, but the language.”
The team also found that fake content on social networks was more frequently shared among people who identify as being autistic.
“In terms of the frequency with which fake news was shared, it’s much higher than what we would expect from people who have autism spectrum,” he said.
The fake content spread like wildfire, he said, but that’s not a surprise given the way people communicate online.
“People who are on social network websites are more likely than people who aren’t to have a good understanding of the science behind social media and how it can be used to propagate misinformation,” Miller explained.
“They’re more likely, for example, to share articles that claim that the moon landings were faked, or to share links to a fake Wikipedia article that claims to show that the Holocaust happened.
They’re not very interested in learning the science.”
Miller said that the study’s findings may also have an impact on online communities where people who might be vulnerable to spreading misinformation are on Facebook.
“There are a lot of communities where, if you’re on Facebook, you’re more open to misinformation, and that could be a potential vulnerability,” Miller added.
Miller said his team plans to expand their algorithm to identify other types of fake news.
“We’ll probably be looking at social network groups, and how those groups might have an influence on the spread of misinformation,” he explained.
Follow NBC Asian America on Facebook and Twitter, and check out the NBC Asian Pacific American YouTube channel for more videos.