Huskers use code to slow spread of fake news

· 3 min read

Huskers use code to slow spread of fake news

Eric Le, Vy Doan
Craig Chandler | University Communication
Eric Le and Vy Doan, students in the department of Computer Science and Engineering, are developing an algorithm that can automatically detect false and misleading information on social media.

As false and misleading stories become more widespread on social media, two Huskers are coding their way to a solution.

Vy Doan and Eric Le, students in the Department of Computer Science and Engineering, are using machine learning to develop an algorithm that can automatically detect fake news on Twitter.

Doan, a junior from Ho Chi Minh City, Vietnam, and Le, a senior from Lincoln, are conducting their research as part of a summer Undergraduate Creative Activities and Research Experience grant under Mohammad Hasan, assistant professor of practice.

"In case you come across a certain set of information that is unfamiliar to you, it's very hard for anyone to know if it's fake news or not, and it's extremely hard for them to take the time to check that," Doan said. "The reason why machine learning is so powerful is that the machine can do this detection a lot faster than a human being."

Machine learning occurs when a computer is introduced to a variety of data points and trained to recognize specific patterns in that data. By analyzing a large amount of tweets from previous years, Doan and Le can find common themes in data — such as location, age and posting frequency of accounts — that indicate a greater likelihood of a tweet being fake.

Developing a successful algorithm, though, is just one piece of the puzzle.

"Even if we are able to figure out what particularly is fake news, then we have to figure out, how do we stop that from propagating through the internet," Le said. "The other part of this is multidisciplinary. It's with computer science, but also has a psychological perspective as well."

Doan and Le want to work towards developing an interface that would warn Twitter users about unreliable sources and prompt them to proceed with caution. The interface could be downloaded as a browser extension, potentially dissuading users to spread false information.

"We choose to believe what we want to believe," Doan said. "So even if the machine can detect fake news, presenting it in a way that would encourage people to look into it more, rather than denying it because it doesn't support their point of view, is a challenge. We try to address both sides to create a more holistic solution to the problem."

Doan, a philosophy double major, and Le, a global studies double major, said working with UCARE has been a fulfilling addition to their computer science education.

"I am interested in seeing technology being applied for social good," Doan said. "The most rewarding part is seeing that the skills I've learned in my degree can be applied to something that can make a better difference in the world."