Modeling the Impact of Group Interactions on Climate-related Opinion Change in Reddit

2025-05-05

AI Summary

Existing opinion dynamics models predominantly rely on pairwise interaction assumptions, failing to capture the prevalent group-level deliberation mechanisms in social media—particularly leading to modeling distortions in climate discourse. To address this, we propose a temporal hypergraph model, the first to integrate hypergraph structures into empirical opinion dynamics research. Leveraging Reddit climate discussion data, our approach explicitly models how multi-user group interactions dynamically shape individual stance evolution. Methodologically, it combines LLM-driven fine-grained comment-level stance annotation, social behavior mining, and a temporal hypergraph neural network. Experiments demonstrate that our model significantly outperforms state-of-the-art pairwise graph-based baselines on individual climate stance prediction. Results confirm that incorporating group interactions yields critical predictive gains for micro-level opinion evolution, establishing a more realistic, structurally grounded paradigm for modeling opinion dynamics on social media.

Technology Category

• Cognitive Modeling & Cognitive Systems: Social Cognition And Interaction
• Data Mining & Knowledge Management: Graph Mining, Social Network Analysis & Community
• Knowledge Representation and Reasoning: Preferences

Application Category

• Social Networks and Social Media: Media and governance, opinion dynamics, filter bubbles, polarization
• Web Mining and Content Analysis: Sentiment analysis and opinion mining
• Graph Algorithms and Modeling for the Web: Foundation models and LLMs for Web-related graphs

Abstract

Opinion dynamics models describe the evolution of behavioral changes within social networks and are essential for informing strategies aimed at fostering positive collective changes, such as climate action initiatives. When applied to social media interactions, these models typically represent social exchanges in a dyadic format to allow for a convenient encoding of interactions into a graph where edges represent the flow of information from one individual to another. However, this structural assumption fails to adequately reflect the nature of group discussions prevalent on many social media platforms. To address this limitation, we present a temporal hypergraph model that effectively captures the group dynamics inherent in conversational threads, and we apply it to discussions about climate change on Reddit. This model predicts temporal shifts in stance towards climate issues at the level of individual users. In contrast to traditional studies in opinion dynamics that typically rely on simulations or limited empirical validation, our approach is tested against a comprehensive ground truth estimated by a large language model at the level of individual user comments. Our findings demonstrate that using hypergraphs to model group interactions yields superior predictions of the microscopic dynamics of opinion formation, compared to state-of-the-art models based on dyadic interactions. Although our research contributes to the understanding of these complex social systems, significant challenges remain in capturing the nuances of how opinions are formed and evolve within online spaces.

Problem

• Modeling group interactions in social media opinion dynamics
• Predicting individual stance shifts on climate change discussions
• Improving opinion formation predictions using temporal hypergraphs

Innovation

• Temporal hypergraph models group dynamics
• Predicts stance shifts in climate discussions
• Uses LLM-estimated ground truth validation