Abstract for: Trust-informed influence: Opinion dynamics in systems with enduring relationships

The world’s communication patterns have changed dramatically with the internet and social media, leading to new patterns of influence and new vulnerabilities. Understanding how these communications change peoples’ opinions is important; typical analyses depend on agent-based opinion dynamics models that do not account for enduring relationships between people. We assert that these enduring relationships are key in understanding opinion dynamics, as they determine trust, which determines how people interpret new information. Here, we introduce the trust-informed influence model, a system dynamics model for simulating opinion dynamics within social systems with enduring relationships. We use this model to evaluate competing influence strategies, and we also introduce a new method that uses reinforcement learning to simulate influence strategies that adapt to new information.