A Queuing Network Model for Eye Movement

Eye movement is a basic human behavior that offers a valuable means to explore human cognitive processes. This article introduces two modeling studies of eye movement. First, random menu search was modeled using a queueing network approach and second, a reinforcement learning algorithm was used to generate various eye movement patterns. Random menu search is a task component involved in many human-machine interfaces and has been modeled with several cognitive models including ACT-R and EPIC. Based upon review of empirical data in menu search, and strengths and limitations of existing models, this article proposes a queueing network model, which has been successfully applied in some other task domains (e.g., response time, driver performance). The queueing network model for random menu search was implemented and evaluated through model simulation. In contrast to existing models that rely on four task-specific strategies to account for data, the queueing network model accounted for the same data using only one strategy already employed in cognitive modeling. To extend this parsimonious, “minimal task strategy” modeling approach, QLearning, one of the reinforcement learning methods, was adopted to generate different patterns in eye movement. The same strategy from random menu search was used to generate eye movement, and the simulated eye movements were qualitatively compared to the human eye movement.