Machine learning of rumour spreading in complex networks

• Grant number: 25/13866-3
• Support Opportunities: Scholarships in Brazil - Scientific Initiation
• Start date: February 01, 2026
• End date: January 31, 2027
• Field of knowledge: Physical Sciences and Mathematics - Mathematics
• Principal Investigator: Francisco Aparecido Rodrigues
• Grantee: Leticia Barbanera de Menezes
• Host Institution: Instituto de Ciências Matemáticas e de Computação (ICMC). Universidade de São Paulo (USP). São Carlos , SP, Brazil

Abstract

This research project aims to investigate how the structural characteristics of the initial spreader node affect the dynamics and outcomes of rumor propagation in complex, non-homogeneous networks. Rumor spreading is modeled as a stochastic process unfolding over directed graphs, where nodes represent individuals and edges capture social interactions. Drawing upon established models from the literature, the study incorporates trust and forgetting mechanisms to more realistically simulate real-world information diffusion. Through extensive simulations on artificially generated scale-free networks, each initiated from a different node, we will examine how topological features such as degree, centrality measures, and clustering coefficients influence key aspects of rumor dynamics-such as cascade size, saturation time, and extinction rates. The collected data will be analyzed using data science and machine learning techniques, including regression models and clustering algorithms, to uncover predictive patterns and infer causal relationships. The findings are expected to enhance our understanding of how information spreads through social systems, with potential applications in marketing, public communication, and misinformation control.