Adversarial Social Influence: Modeling Persuasion in Contested Social Networks

2025-10-01

AI Summary

This paper addresses adversarial persuasion modeling in multi-competitor social networks, focusing on scalable and interpretable identification of network influence levers. We propose the Social Influence Game (SIG) framework, which formulates multi-player adversarial influence as a difference-of-convex (DC) programming problem for the first time, grounded in DeGroot opinion dynamics. Theoretically, we establish an asymptotic analysis foundation for large-scale networks. Algorithmically, we design an Iterative Linearization (IL) solver to efficiently approximate Nash equilibrium strategies while preserving interpretability. Experiments demonstrate that the IL solver incurs <7% solution quality loss compared to exact methods, achieves over 10× speedup, and scales to million-node networks on both synthetic and real-world graphs. Our core contribution is a novel adversarial influence modeling paradigm that jointly ensures scalability, interpretability, and theoretical rigor.

Technology Category

• Game Theory and Economic Paradigms: Adversarial Learning
• Multiagent Systems: Adversarial Agents
• Search and Optimization: Adversarial Search

Application Category

• Social Networks and Social Media: Influence propagation, information diffusion, and the prediction on networks
• Graph Algorithms and Modeling for the Web: Efficient manipulation of static and dynamic Web-related graphs
• Economics, Online Markets and Human Computation: Social networks and social learning

Abstract

We present the Social Influence Game (SIG), a framework for modeling adversarial persuasion in social networks with an arbitrary number of competing players. Our goal is to provide a tractable and interpretable model of contested influence that scales to large systems while capturing the structural leverage points of networks. Each player allocates influence from a fixed budget to steer opinions that evolve under DeGroot dynamics, and we prove that the resulting optimization problem is a difference-of-convex program. To enable scalability, we develop an Iterated Linear (IL) solver that approximates player objectives with linear programs. In experiments on random and archetypical networks, IL achieves solutions within 7% of nonlinear solvers while being over 10x faster, scaling to large social networks. This paper lays a foundation for asymptotic analysis of contested influence in complex networks.

Problem

• Modeling adversarial persuasion in contested social networks
• Optimizing influence allocation under opinion dynamics
• Developing scalable algorithms for large network analysis

Innovation

• Modeling adversarial persuasion via Social Influence Game
• Proving influence optimization as difference-of-convex program
• Developing scalable Iterated Linear solver for networks