Agent-Based Models in Social Physics

TL;DR

This paper reviews agent-based models in social physics, highlighting their structure, behavior, and applications across various social phenomena, emphasizing their ability to simulate complex emergent behaviors.

Contribution

It provides a comprehensive overview of ABM in social physics, including models, platforms, and applications in econophysics, crowd dynamics, and opinion formation.

Findings

ABM captures emergent social phenomena.

Different platforms have unique advantages and disadvantages.

Models like Sugarscape and minority/majority games are key examples.

Abstract

We review the agent-based models (ABM) on social physics including econophysics. The ABM consists of agent, system space, and external environment. The agent is autonomous and decides his/her behavior by interacting with the neighbors or the external environment with the rules of behavior. Agents are irrational because they have only limited information when they make decisions. They adapt using learning from past memories. Agents have various attributes and are heterogeneous. ABM is a non-equilibrium complex system that exhibits various emergence phenomena. The social complexity ABM describes human behavioral characteristics. In ABMs of econophysics, we introduce the Sugarscape model and the artificial market models. We review minority games and majority games in ABMs of game theory. Social flow ABM introduces crowding, evacuation, traffic congestion, and pedestrian dynamics. We also review ABM for opinion dynamics and voter model. We discuss features and advantages and disadvantages of Netlogo, Repast, Swarm, and Mason, which are representative platforms for implementing ABM.