What is it about?
Disinformation, spreads rapidly on social media and can harm public trust, democratic systems and social stability. In our work, we developed a model to understand how disinformation spreads in online networks and what makes people adopt, reject, or ignore such narratives. We used a physics-based approach called mean-field theory to simplify complex interactions between users. Our model considers how people are influenced by others, how they remember past beliefs and how large-scale campaigns (such as disinformation pushes or fact-checking efforts) affect opinions over time. We discovered that belief shifts often depend on "tipping points." If influence or pressure crosses a certain threshold, disinformation can spread quickly and become difficult to reverse. Our findings shed light on when societies are most vulnerable and offer guidance on how to intervene early and effectively to reduce the harm caused by false information.
Featured Image
Photo by Hartono Creative Studio on Unsplash
Why is it important?
Our work is unique and timely because it applies a powerful physics-based framework to capture the real-world complexity of how disinformation spreads, including memory effects, varying influence levels, and external campaigns. Unlike earlier models, we identify tipping points and conditions where disinformation becomes self-sustaining. This helps policymakers and platforms act early, making the work highly relevant in today's fight against online manipulation.
Perspectives
This publication is important because it offers a novel and rigorous way to understand how disinformation spreads and takes root in society, something that traditional detection tools often fail to explain. Modeling social influence, memory, and external campaigns within a unified framework allows this work to move beyond the limits of content analysis and basic network models. It captures the dynamic, often nonlinear nature of belief formation, offering a deeper understanding of how disinformation becomes entrenched. This makes the research especially valuable for designing timely, targeted interventions that can reduce the real-world impact of false narratives.
Spyridon Evangelatos
Netcompany SA
Read the Original
This page is a summary of: Modeling Disinformation Spread in Social Networks: Phase Transitions and Mean-Field Analysis, ACM Transactions on the Web, October 2025, ACM (Association for Computing Machinery),
DOI: 10.1145/3747287.
You can read the full text:
Contributors
The following have contributed to this page
