From pretraining data to language models to downstream tasks: Tracking the trails of political biases leading to unfair NLP models

S Feng, CY Park, Y Liu, Y Tsvetkov - arXiv preprint arXiv:2305.08283, 2023 - arxiv.org
Language models (LMs) are pretrained on diverse data sources, including news, discussion
forums, books, and online encyclopedias. A significant portion of this data includes opinions …

Botmoe: Twitter bot detection with community-aware mixtures of modal-specific experts

Y Liu, Z Tan, H Wang, S Feng, Q Zheng… - Proceedings of the 46th …, 2023 - dl.acm.org
Twitter bot detection has become a crucial task in efforts to combat online misinformation,
mitigate election interference, and curb malicious propaganda. However, advanced Twitter …

Large Language Models Help Humans Verify Truthfulness--Except When They Are Convincingly Wrong

C Si, N Goyal, ST Wu, C Zhao, S Feng… - arXiv preprint arXiv …, 2023 - arxiv.org
Large Language Models (LLMs) are increasingly used for accessing information on the web.
Their truthfulness and factuality are thus of great interest. To help users make the right …

Identifying Search Directives on Social Media

RE Robertson, A Dunphy, S Grossman… - Journal of Online Trust …, 2023 - tsjournal.org
This study introduces methods for identifying search directives—content that could prompt
an online search—and explores their presence on social media. Search directives can be …

Botpercent: Estimating bot populations in twitter communities

Z Tan, S Feng, M Sclar, H Wan, M Luo, Y Choi… - arXiv preprint arXiv …, 2023 - arxiv.org
Twitter bot detection is vital in combating misinformation and safeguarding the integrity of
social media discourse. While malicious bots are becoming more and more sophisticated …

Designing Gig Worker Sousveillance Tools

K Do, M De Los Santos, M Muller… - Proceedings of the CHI …, 2024 - dl.acm.org
As independently-contracted employees, gig workers disproportionately suffer the
consequences of workplace surveillance, which include increased pressures to work …

OSINT Research Studios: A Flexible Crowdsourcing Framework to Scale Up Open Source Intelligence Investigations

A Mukhopadhyay, S Venkatagiri, K Luther - Proceedings of the ACM on …, 2024 - dl.acm.org
Open Source Intelligence (OSINT) investigations, which rely entirely on publicly available
data such as social media, play an increasingly important role in solving crimes and holding …

Design and analysis of tweet-based election models for the 2021 Mexican legislative election

A Vigna-Gómez, J Murillo, M Ramirez, A Borbolla… - EPJ Data …, 2023 - epjds.epj.org
Modelling and forecasting real-life human behaviour using online social media is an active
endeavour of interest in politics, government, academia, and industry. Since its creation in …

Inclusive Portraits: Race-Aware Human-in-the-Loop Technology

C Flores-Saviaga, C Curtis, S Savage - … of the 3rd ACM Conference on …, 2023 - dl.acm.org
AI has revolutionized the processing of various services, including the automatic facial
verification of people. Automated approaches have demonstrated their speed and efficiency …

GigSense: An LLM-Infused Tool forWorkers' Collective Intelligence

K Imteyaz, C Flores-Saviaga, S Savage - arXiv preprint arXiv:2405.02528, 2024 - arxiv.org
Collective intelligence among gig workers yields considerable advantages, including
improved information exchange, deeper social bonds, and stronger advocacy for better …