Dr. Kiril Avramov, an ISP Fellow, and the team at the UT Global Disinformation Lab co-authored a report on “Identifying Disinformation Using Rhetorical Devices in Natural Language Models” with researchers at Sandia National Laboratories and North Carolina State University. Through analysis of disinformation and neutral articles, the initial findings demonstrated significantly more occurrences of certain rhetorical devices, or methods of persuasive speaking or writing, in disinformation compared to neutral counterparts. The results were a promising proof of concept allowing GDIL to move forward with developing a machine-learning model to recognize rhetorical devices and automatically detect disinformation. Read the full report HERE.