TAMPA, Florida. (February 3, 2022) – Social media sites continue to amplify misinformation and conspiracy theories. To address this concern, an interdisciplinary team of computer scientists, physicists, and social scientists led by the University of South Florida (USF) came up with a solution to ensure that social media users are exposed to more reliable sources of information.
In their study published in the journal Nature Human behavior, the researchers focused on the recommendation algorithm used by social media platforms to prioritize content displayed to users. Rather than measuring engagement based on the number of users and page views, the researchers looked at what content gets amplified on a News Feed, focusing on a news source’s Trust Score and the political diversity of its audience.
“Low-quality content is engaging because it’s consistent with what we already know and love, whether it’s accurate or not,” said Giovanni Luca Ciampaglia, assistant professor of computer science and engineering at USF. “As a result, misinformation and conspiracy theories often go viral among like-minded audiences. The algorithm ends up picking the wrong signal and continues to promote it. To break this cycle, look for content that is engaging, but aimed at a diverse audience, not a like-minded audience.
Working with researchers from Indiana University and Dartmouth College, the team created a new algorithm using web traffic and self-reported partisanship data from 6,890 people that reflects America’s diversity in gender, race and political affiliation. The data was provided by online polling company YouGov. They also looked at the trust scores of 3,765 news sources based on the NewGuard Trust Index, which rates news sources on several journalistic criteria, such as editorial responsibility, accountability and financial transparency. .
They found that incorporating the partisan diversity of a news audience can increase the reliability of recommended sources while providing users with relevant recommendations. Since the algorithm is not exclusively based on engagement or popularity, it is still able to promote trusted sources regardless of their partisanship.
“This is particularly welcome news for social media platforms, especially as they have been reluctant to introduce changes to their algorithms for fear of criticism over partisan bias,” said co-author Filippo Menczer, Distinguished Professor of Computer Science and Luddy Computer Science at Indiana University. .
The researchers say that platforms could easily include audience diversity in their own recommendation algorithms because diversity metrics can be derived from engagement data, and platforms already log this kind of data whenever users click “Like” or share something on a newsfeed. Ciampaglia and his colleagues propose that social media platforms adopt this new strategy in order to help prevent the spread of misinformation.
About the University of South Florida
The University of South Florida, a high-impact global research university dedicated to student success, generates an annual economic impact of more than $6 billion. Over the past 10 years, no other public university in the nation has moved up the US News and World Report national university rankings faster than USF. Serving more than 50,000 students at campuses in Tampa, St. Petersburg, and Sarasota-Manatee, USF is designated a preeminent state research university by the Florida Board of Governors, placing it in the category of most elite among the 12 public universities in the state. USF has gained wide national recognition for its success in graduating underrepresented minority and low-income students at rates equal to or better than white and higher-income students. USF is a member of the American Athletic Conference. Learn more at www.usf.edu.
Nature Human behavior
The title of the article
Diversity of political audience and reliability of information in algorithmic ranking
Publication date of articles
Warning: AAAS and EurekAlert! are not responsible for the accuracy of press releases posted on EurekAlert! by contributing institutions or for the use of any information through the EurekAlert system.