A Micro-Level Data-Calibrated Agent-Based Model: The Synergy between Microsimulation and Agent-Based Modeling.

Area of Science:

• Computational science
• Artificial life research
• Social simulation

Background:

• Artificial life (ALife) studies natural life processes and evolution through computational models, robotics, and biochemistry.
• Existing ALife research often requires specialized expertise in parallel and distributed computing.
• A need exists for accessible frameworks supporting ALife experiments in computational modeling.

Purpose of the Study:

• To develop a user-friendly, parallel, and distributed agent-based modeling environment for Artificial Life (ALife) research.
• To implement a hybrid model combining microsimulation and agent-based modeling for artificial society generation.
• To analyze population dynamics and estimate policy costs using simulated artificial societies.

Main Methods:

• Designed and built a parallel and distributed agent-based modeling framework requiring no specialized computing expertise.
• Implemented a hybrid model integrating microsimulation (real data-driven behavior) and agent-based modeling (agent interaction).
• Utilized Korean population census data to inform agent behaviors within the artificial society.

Main Results:

• Successfully generated an artificial society capable of simulating population dynamics.
• Analyzed social scenarios and agent interactions to understand population trends.
• Estimated future pension policy costs based on projected population structures from the artificial society.

Conclusions:

• The proposed ALife framework and hybrid model effectively support the simulation of complex social dynamics.
• ALife techniques offer valuable tools for analyzing social issues and informing policy decisions.
• The framework democratizes advanced computational modeling for a broader range of scientists and researchers.