Regensburg 2025 – scientific programme
Parts | Days | Selection | Search | Updates | Downloads | Help
SOE: Fachverband Physik sozio-ökonomischer Systeme
SOE 10: Focus Session: Large Language Models, Social Dynamics, and Assessment of Complex Systems
SOE 10.8: Talk
Thursday, March 20, 2025, 17:15–17:30, H45
Understanding Information Spread in Social Networks through the Lense of Self-Organized Criticality: A Study on Telegram — •Roman Ventzke1,2, Anastasia Golovin1,2, Sebastian Bernd Mohr1,2, Andreas Schneider1,2, and Viola Priesemann1,2 — 1Max-Planck-Institut für Dynamik und Selbstorganisation, Göttingen — 2Georg-August-Universität Göttingen
To effectively address the proliferation of misinformation in social media, one fundamentally needs to understand how information spreads in online social networks. We investigate the dynamics of information spread using a large dataset from the messenger platform Telegram, showing that information spread in the networks happens in bursty avalanches with scale-free statistics that resembles critical behavior of physical systems.
We find that the critcal exponents of the system can be well described by a mean-field Random Field Ising Model (RFIM), alluding to an important role of complex contagion and peer-pressure effects in the propagation of information. By coarse-graining dynamics in the topic space we show additional evidence that the process indeed belongs to the RFIM class. We demonstrate through simulations that the spreading process on the network can be well-described by mean-field models and discuss how self-regulation of the network gives rise to criticality.
Keywords: social media; self-organized criticality; random field ising model; complex contagion; networks