CS Research Talk: Computational Models of Persuasive Media

Dr. Rebecca Hwa
University of Pittsburgh
Professor of Computer Science

Friday, October 15, 2021 |11:30a.m. – 12:30 p.m.
Join via Zoom
or 
In Person @
800 22nd Street NW, Washington, DC, 20052
SEH Lehman Auditorium (B1220)

Computational Models of Persuasive Media

Abstract

Persuasive language presents several unique challenges to the automatic processing of text. Common rhetorical devices used for persuasion are non-literal and implicit, so systems need to be able to disambiguate metaphoric usages and recognize the underlying intentions in text. Moreover, human interpretations of the underlying message may be subjective as well, so it is often difficult to obtain “ground truths” without introducing unintended biases. These characteristics also limit the availability of appropriate training data needed for machine learning approaches. This talk highlights some of our recent efforts in addressing these challenges. I will begin by describing our work on distinguishing between figurative and literal usages of idiomatic expressions in language. I will then show how the ideas extend to other domains such as understanding advertisements, which involves both text and images. Another domain of persuasion that we have examined is social media  about controversial topics (e.g., gun control). I will present some studies on the impact of implicit political ideology biases and methods to mitigate them. I will also be talking about some semi-supervised approaches that we’ve been exploring in order to cope with the training data limitations in these domains.

Bio

Rebecca Hwa is Professor of Computer Science at the University of Pittsburgh. She received her BS in Computer Science and Engineering from UCLA in 1993, and her PhD in Computer Science from Harvard University in 2001. Her recent research focuses on understanding persuasion from a computational linguistics perspective. Some of her recent projects include: modeling student behaviors in revising argumentative essays, identifying symbolisms in visual rhetorics, and recognizing group biases in social media. Dr Hwa is a recipient of the NSF CAREER Award. Her work has also been supported by NIH and DARPA. She is currently serving as a Program Officer in the Computer and Information Science and Engineering directorate at NSF.