Analysis of Opinion Spread through Migration and Adoption in Agent Communities

Agent based simulation can be an important tool for enhancing our understanding of the dynamics of complex systems. This paper studies the migration and behavior adoption patterns of agents in a system of agents situated in geographically distributed communities. We consider two agent types with binary and continuous states. Agents either probabilistically adopt the predominant state in their community or migrate to another community more consistent with their state. This adoption and migration model is specified with few parameters but can support a wide range of interesting behaviors and emergent phenomena. We investigate the conditions for the convergence of communities to different states and the role of migration and adoption on convergence. We experiment with different migration probability distributions and adoption rates and analyze the resulting patterns of population and state distributions.