Leveraging the learning process in health through clinical cases simulator

This paper presents a multi-agent learning system for health care practitioners named SimDeCS (Simulation for Decision Making in the Health Care Service). The main contribution of this work is the system architecture, model-learning environment supported by artificial intelligence techniques, and its evaluation as a educational software. The SimDeCS was designed as a multi-agent system, composed by three intelligent agents: Domain Agent, Learning Agent and Mediator Agent. The Domain Agent implements the knowledge model by probabilistic reasoning (Bayesian networks), with the knowledge encoded by human experts. The pedagogical strategies emerge from an influence diagram, based on the student's conduct during the simulation. Some results related with the SimDeCs evaluation are presented.