Publications

Exposure to news and political content on TikTok: A Case Study Around the 2024 US Election

Published in Accepted at FAccT 2026, 2026

TikTok is a major social media platform with a distinctive centralized recommendation system used by 56% of adults under 34 years of age in the U.S. In this study, we deploy approximately 200 sock puppet accounts during the 2024 U.S. election to examine how TikTok’s recommendation algorithm might respond to users’ interests in news and political content. Drawing on over 217,064 unique videos viewed by these accounts, we provide a data-driven analysis of the extent to which accounts with different interests are exposed to news media, and election-related political content. We find that TikTok rarely shows news or political content to accounts. Even the accounts that only engage with news and political content are recommended less than 0.30 total news videos per browsing session on average, though they are recommended slightly more non-news political content (1.60 total political videos on average). We find that users are recommended far more election-related political content posted by everyday users than by news organizations. The lack of news content on TikTok, especially for users who have demonstrated an interest in news, has implications for political knowledge amongst social media users.

YouTube content on wildlife engages audiences but rarely drives meaningful conservation action

Published in Coummunications Sustainability, 2026

Here, we analyze wildlife-related YouTube videos to assess dominant themes, audience engagement, and the frequency of conservation-related calls to action.

Recommended citation: Van Berkel, D., Gautam, N., Carter, N.H., Di Minin, E., Zhang, Y., Mei, H., Yin, S. & Tomkins, S. (2026) " YouTube content on wildlife engages audiences but rarely drives meaningful conservation action." Communications Sustainability. 1(1).
Download Paper

What does the public want local government to hear?

Published in Journal of Quantitative Description: Digital Media, 2026

In this paper we analyze a large set of public comments across the state of Michigan.

Recommended citation: Ge, C., Zhang, J., Xu, H., Krupnikov, Y., Bednar, J., & Tomkins, S. (2025) " What does the public want their local government to hear? A data-driven case study of public comments across the state of Michigan." Journal of Quantitative Description: Digital Media,. 5.
Download Paper

What Do You Learn About Hard News from Soft News Outlets - A Case Study from People Magazine Online

Published in AAAI Conference on Web and Social Media, 2025

As people increasingly avoid outlets with hard news, our work shows that they may still encounter political information through soft news outlets, a finding that has critical implications for political communication.

Recommended citation: Yan, J., Milita, K., Krupnikov, Y., & Tomkins, S. (2025) " What Do You Learn About Hard News from Soft News Outlets - A Case Study from People Magazine Online." Proceedings of the International AAAI Conference on Web and Social Medi. 19(1).
Download Paper

PUBLICSPEAK: Hearing the Public with a Probabilistic Framework

Published in AAAI Technical Track on AI for Social Impact Track, 2025

Our method for classifying public remarks in city council meetings.

Recommended citation: Xu, T., Brown, E.M., Dwyer,D., & Tomkins, S. (2025) " PUBLICSPEAK: Hearing the Public with a Probabilistic Framework." Proceedings of the AAAI Conference on Artificial Intelligence. 39(27).
Download Paper