Cognitive Modeling as a Method for Agent Development in Artificial Intelligence

This research aims to expand the applications of cognitive modeling by exploring the use of a testbed approach to modeling human behaviour. Newell's complex task analysis method (1990) was applied to model an agent completing a complex task through incremental and iterative design to test the agent's ability to fluidly monitor and react to internal and external interruptions and perform the tasks at an expert level successfully through simulated challenge sets within different environments. A single agent is developed incrementally over five stages of distinct simulated challenge sets, with testing for backwards compatibility. The testbed approach and incremental development provides insight to agent-specific cognitive structure functionality, and to the process of combining microcognition and macrocognition in cognitive modeling.