Newhouse professors receive funding for work on technology to combat fake news

By Wendy S. Loughlin

July 28, 2020

Stephen Masiclat and Regina Luttrell will work to refine a theoretical framework for the creation and testing of AI algorithms that can identify manipulated media

Two professors from Syracuse University’s S.I. Newhouse School of Public Communications have received a $830,958 subcontract agreement for the development of technology to detect manipulated media and combat the spread of fake news.

Stephen Masiclat and Regina Luttrell
Stephen Masiclat and Regina Luttrell

Stephen Masiclat, professor and director of new media management and director of the Thomas and Lisa Mandel Experimental Media Lab, and Regina Luttrell, assistant professor of public relations and director of the W2O Emerging Insights Lab, will work to refine a theoretical framework for the creation and testing of AI algorithms that can identify manipulated media. They will collaborate with researchers from private industry and academia.

The 48-month subcontract is part of the Semantic Forensics (SemaFor) program, funded by an $11.9 million Defense Advanced Research Projects Agency contract with PAR Government Systems Corp. The program seeks to create a system for automatic detection, attribution and characterization of falsified media assets.

“The challenge of fake news and disinformation is something we as communications educators have an obligation to address,” says Newhouse Dean Mark J. Lodato. “This is a new area of research for the Newhouse School, and allows us to contribute to the ongoing national conversation about the importance of reliable, fact-based information for the health of our society. Steve and Gina are doing important work.”

Masiclat and Luttrell have already built a preliminary theoretical framework consisting of an eight-dimensional analysis, and will work with researchers to test and refine their ideas.

They will create large data sets—a massive archive of both real and fake news stories based on their theoretical framework—that will train and validate AI algorithms. “Over the next four years we will evaluate various aspects of our proposed ‘theory of semantic consistency’ that can be used to create and test AI algorithms that detect key flaws,” Masiclat says. “This will allow us to develop a method for separating reliable journalism from deliberate misinformation.”

Wider access to automated manipulation technologies, coupled with the ease of sharing provided by social media platforms, has increased the threat posed by manipulated media, according to DARPA.

“This work signals Newhouse’s commitment to preserving the First Amendment and addressing the impact misinformation spread across social media has on society,” Luttrell says. “The proliferation of fake news over the past few years has caused numerous problems. Grappling with the many questions plaguing the role of truth and trust in news media, social media and society is paramount. It’s our intention that this research will help detect and combat disinformation.”

Masiclat and Luttrell will form a team of doctoral student researchers, based in Newhouse’s Experimental Media and W2O Emerging Insights Labs, to assist with the research.