Collective Coordination through Online Social Media

ABSTRACT

Grassroots coordination through the Social Web holds promises for addressing global challenges, such as climate change, pandemics, and mass migration crises. However, the understanding of cooperation in online communities is hindered by restrictive data collection policies and by the limited ability of traditional computational analysis tools to grasp the complexity of social behavior.

To overcome these challenges, our research applies advanced Natural Language Processing methods to social media conversations to explore how social science dimensions correlate with cooperation outcomes in social networks. We delve into the dynamics of platforms like YouTube and TikTok, investigating how content creators can strategically leverage their differences to guide aligned responses within their audience. In particular, we study how the public’s reactions to carefully crafted narratives have the potential to foster collective efforts.
In parallel, our research delves into Large Language Models (LLMs), assessing their efficacy in classifying and generating synthetic data for low-resource social science domains. Further exploration involves LLM-based agents in controlled persuasion dialogue settings related to climate change discourse, where distinct social personalities are assigned. The goal of our project is to establish a platform where both human and LLM agents can interact, leveraging insights from social media dynamics to facilitate cooperative discourse. This collaborative effort aims to address global challenges through the coordinated endeavors of human and artificial intelligence.

BIOS

Arianna Pera is a PhD Student at the IT University of Copenhagen, working within the Computer Science Department and the NEtwoRks, Data and Society (NERDS) Research Group. Her primary research interest is in the intersection between Network Science and NLP, focusing on the interplay between online social networks and linguistic interactions in determining social coordination, especially in the context of climate change mitigation. She is part of the COllective COordination through Online Social Media (COCOONS) project, under the supervision of Luca Maria Aiello. Before joining ITU, she received a B.Sc. in Statistics and Information Management and an M.Sc. in Data Science from the University of Milano-Bicocca.

Anders Giovanni Møller is a PhD student at the IT University of Copenhagen, specializing in data science with a particular focus on natural language processing and generative AI. Holding a Master’s degree in Data Science from the same institution, Anders has carved a niche in the application of AI and large language models within various contexts. His current research investigates the integration of LLMs in online social platforms, exploring how generative AI can enhance and influence digital interactions and community dynamics.

He is working in the NEtwoRks, Data and Society (NERDS) Research Group, and is also a part of the COllective COordination through Online Social Media (COCOONS) project supervised by Luca Maria Aiello.