The recent publications in the field of computational social science and digital communication highlight a significant shift towards understanding and mitigating the effects of digital platforms on societal polarization, misinformation, and collective action. Researchers are increasingly leveraging advanced machine learning techniques, including transformer models and large language models (LLMs), to analyze social media content, detect extremist traits, and predict social support levels. These studies underscore the importance of platform dynamics in shaping online discourse, with particular attention to how search engines and social media platforms curate information and influence public opinion. Additionally, there is a growing focus on the role of social media in mobilizing collective action and the spread of misinformation, with innovative methodologies being developed to track news narratives and quantify political polarization. The field is also exploring the behavioral differences in information operations across linguistic and cultural contexts, highlighting the need for adaptable detection methods. Furthermore, the impact of AI on local communities and the governance of AI technologies are emerging as critical areas of research, reflecting the broader societal implications of technological advancements.

Noteworthy Papers

• Advanced Machine Learning Techniques for Social Support Detection on Social Media: Introduces a novel approach using transformers and zero-shot learning for classifying social support content, achieving significant improvements in macro F1 scores.
• Search engines in polarized media environment: Audits Google and Bing's curation of election-related information, revealing a tendency to prioritize left-leaning media sources and highlighting the role of search engines in mirroring partisan divides.
• Unifying the Extremes: Proposes a generalized psychosocial model of extremism, demonstrating its effectiveness in predicting user radicalization with high accuracy.
• Using LLMs to Infer Non-Binary COVID-19 Sentiments of Chinese Micro-bloggers: Utilizes Llama 3 8B for sentiment analysis on Weibo, offering insights into public opinion shifts during the COVID-19 crisis.
• Empirical Power Analysis of a Statistical Test to Quantify Gerrymandering: Computationally verifies the power of statistical tests used in gerrymandering cases, emphasizing the importance of bias metric selection.
• Expressing One's Identity Online: Maps socio-political identity expression across EU countries, identifying divisive topics and partisan differences in online self-presentation.
• Polarized Patterns of Language Toxicity and Sentiment of Debunking Posts on Social Media: Investigates the relationship between language toxicity and social polarization in debunking efforts, highlighting platform-specific dynamics.
• Unveiling Voices: A Co-Hashtag Analysis of TikTok Discourse on the 2023 Israel-Palestine Crisis: Explores activism and propaganda on TikTok, using network analysis to categorize discourse on the Israel-Palestine conflict.
• Extracting Participation in Collective Action from Social Media: Develops a novel suite of text classifiers to identify and categorize participation in collective action, contributing to computational social science research.
• Quantifying Polarization: Evaluates polarization measures and introduces a novel adaptation of Kleinberg's burst detection algorithm for improved mode detection in polarized distributions.
• Unveiling Behavioral Differences in Bilingual Information Operations: Analyzes behavioral differences in information operations across English- and Spanish-speaking communities, emphasizing the need for culturally adaptable detection methods.
• DomainDemo: Introduces a unique dataset linking domains shared on Twitter with user demographics, offering insights into political discourse among different sociodemographic groups.
• Tracking the Takes and Trajectories of English-Language News Narratives: Utilizes large language models to track news narratives across thousands of websites, identifying influential sources in spreading propaganda.
• The Spread of Virtual Gifting in Live Streaming: Examines the spread of gifting behavior on Twitch, providing evidence on the factors influencing online prosocial behavior.
• Local US officials' views on the impacts and governance of AI: Surveys local policymakers on AI's societal impacts and regulatory attitudes, highlighting the need for capacity-building and bipartisan coordination in AI governance.