Modeling Effects of Behavior Moderators on Performance: Evaluation of the MAMID Methodology and Architecture
Eva Hudlická · 2015
The past five years have witnessed a growth of interest in modeling behavior moderators within cognitive architectures. Incorporating these effects within human behavior models driving training simulations is essential for producing more realistic simulated agent performance and improved training. We describe a methodology for modeling a broad range of interacting behavior moderators in terms of architecture parameters, and a cognitive architecture that implements this methodology. We then present results of an evaluation experiment that demonstrates the architecture’s ability to produce observable behavior differences resulting from distinct individual profiles. The demonstration illustrates distinct behaviors of three types of commanders (‘normal’, ‘anxious’, and ‘aggressive’) within a simulated Stability and Support Operations scenario.