Fuzzy Cognitive Maps in Agent Based Models: A Practicial Implementation Example

Abstract

Agent Based Social Simulation (ABSS) has been used to create simulations with agents that contain cognitive models to achieve more realistic and explainable results by mimicking human cognition. However when using this technique the decision model used by agents is often a black box. Fuzzy Cognitive Maps (FCMs) have been used to provide clear and transparent representations of cognition and could be used as models for agents in simulations to achieve results with better transparency. Utilizing FCMs to represent agents' cognitions in social and organizational modeling is still an emerging field with limited work done to show how FCM and ABM can be combined for greater model interpretability. This paper provides 2 major contributions: (1) the use of FCM for agent's design in social simulation as compared to previous research using different cognitive models, and (2) a system architecture consisting of 8 components that combine FCM and ABM for simulation.