Tuesday, April 30, 2024, 9am

Before the ubiquity of social media, the information space was dominated by a small number of trusted originators (e.g., news organizations). This paradigm was shattered and replaced with a more diffused information environment where content originates from often unknown actors and is propagated nearly instantaneously. The need to rapidly identify factually inaccurate information (misinformation) in this environment is critical. Though some solutions exist to this problem, most current systems rely on human-driven fact-checking.

Existing fact-checking systems cannot operate at scale and are subject to significant human bias. In this thesis, I will describe a methodology for a fully automated misinformation detection pipeline that operates on multiple social media platforms. The pipeline leverages natural language processing (NLP) approaches to find, extract, and contextualize claims that could contain misinformation. Another set of NLP models and network science approaches assigns a likelihood to the truth value of these claims to arrive at a final misinformation likelihood score.

As part of this thesis, I will validate the misinformation detection pipeline in terms of its ability to accurately detect misinformation in labeled datasets as well as its research utility in the social cybersecurity domain by applying it in case studies on multiple social media platforms that focus on diverse topics/communities. The resulting system will serve as a valuable part of the social cybersecurity researchers' toolkit, to be used alongside the BEND framework to characterize the information environment and ultimately inform effective countermeasures.

Thesis Committee
Kathleen Carley (Chair)
Brandy Aven
Hong Shen
COL David Beskow (United States Military Academy)

Additional Information

In Person and Zoom Participation. See announcement.

Event Type: Thesis Proposals
Room Number: In Person and Virtual - ET
Building: TCS Hall 460 and Zoom
Speaker's Name: IAN KLOO
Speaker Website: sc.s3d.cmu.edu…
Speaker's Professional Title: Ph.D. Student, Ph.D. Program Societal Computing, Software and Societal Systems Department, Carnegie Mellon University
Talk Title: Automated Misinformation Detection with Natural Language Processing and Network Models
Event Poster Title: Poster
Event Poster URL: www.cs.cmu.edu…
For More Information: cherold@cs.cmu.edu | aroudebu@andrew.cmu.edu
Affiliations: Software and Societal Systems Department (S3D)
Organization(s): School of Computer Science