Evolving multi-agent network structure with organizational learning

Organizational structure changes over time due to various reasons, such as organizational learning, situation changes and personnel turnovers, etc. Estimating the structure changes will reflect organizational performance changes, emergent leaders, new key links, etc. This paper introduces a multi-agent model that simulates the organizational structure evolution over time. The simulated structure evolution will be driven by the organizational learning procedure that we devised. We perform virtual experiments with two distinct cases, an organization with the learning mechanism and the other one without learning. The performances of the two case organizations were examines under situation change assumptions. The organization with learning mechanism was better than the other when situation changes were predictable. We also scan the network topology changes over time, and we identified that the average distance among the nodes gets smaller as learning proceeds. This work is a preliminary effort to examine the effect of organizational learning and to formulate the evolution of organizational structures.