I am the CCI Board of Visitors Assistant Professor in the School of Information Sciences in the College of Communication and Information at The University of Tennessee Knoxville. I am also a Joint Faculty Assistant Professor in The Bredesen Center's Data Science and Engineering program, a Founding Fellow of the CCI Information Integrity Institute, an Affiliate of the Center for National Security and Foreign Affairs at the Baker School for Public Policy and Public Affairs, and co-director of The MeLa Lab, a multi-university lab that works on problems related to online news and media.
I am a highly interdisciplinary, computational social scientist who takes a pragmatic approach to research. I study the production of, consumption of, and interventions to misinformation, disinformation, and malign influence in media ecosystems.
This work fits into two related categories:
1. Content moderation and information interventions, ranging from automated detection techniques (or alternatives to automated techniques) to content warning label designs. When I build or study tools, I prioritize ethics, safety, and practicality over building narrowly accurate tools. When I design intervention experiments, I typically focus on the use of heuristics in information trust decisions.
Examples: ICWSM25, BIT23, CACM22, ICWSM19a
2. Media ecosystems, where media can include news media, social media influencers, and whole social media platforms. Importantly, I am not only referring to media consumption but also media production. Within this area, I have particular expertise in online news media.
Examples: ICWSM22, ICWSM19b, Websci22, ICWSM24
In addition to this core work, I collaborate with social scientists to study the relationships between and influences of culture, identity, and lived experiences on perceptions, with the long-term goal of using what is learned from this work to inform my work on media ecosystems and interventions.
Examples: HSSC23, SNSS24, CRESP23, AB23
Across all of these categories, I tend to focus on trust as the key metric (trust in media, trust in content moderation, trust in automated systems, etc.). I like to think about this work within the pipeline shown in the Figure below, where I study various facets of media, the humans who consume it, and interventions in between.
My background is in Computer Science. Hence, methodologically, I am most experienced in using computational methods from natural language processing, network science, and machine learning on digital trace data. However, I often design human experiments and employ qualitative methods on both digital trace data and survey data to aid my research.
Outside of academia, I spend much of my time collecting sports cards, vintage programs, and memorabilia. After many years of collecting, I have particular expertise in vintage rookie cards and autographs from hall of famers in football, basketball, and hockey. What I most enjoy about this hobby is that sports cards – like sports media - can uniquely capture moments in time and culture, when thought about within their historical context. I sometimes talk about this collection on Instagram @TheProfessorsCards.
Here is a Pecha-Kucha talk (20 slides x 20 seconds) I gave at UT Mic/Nite in 2024 on if we should let AI tell us what is true:
As a brief sample of my work in automated news veracity detection, check out a talk I gave at ICML 2022:
Email: bhorne6 at utk dot edu
Google Scholar: Scholar Profile
Github: benjamindhorne
Office: 440B Communications Building, 1345 Circle Park, Knoxville TN 37996
Erdős Number: 3 (Erdős -> Goldberg -> Adalı -> Horne)