Role Explication of Simulation in Intelligent Training Systems by Training Task Ontology ˚Lai JIN, Mitsuru IKEDA, Riichiro MIZOGUCHI ISIR, Osaka University Email:{kin,ikeda,miz}@ei.sanken.osaka-u.ac.jp 8-1, Mihogaoka, Ibaraki, 567, Osaka, JAPAN Yoshiyuki TAKAOKA TOKO SEIKI Company Email:charly@kodama.cdd.toko-s.co.jp 3-14-40, Senrioka, Setushi, 566, Osaka, JAPAN 1 Introduction With the advancement in technology, computers have been used for training novice operators, for it is more cost-effective than the real system, and with the recent development of Multi-Media technology, a simulator environment can be realized in computer just like a real-time control system. But on the other hand, the more complex and the more advanced the technology is, the more difficult to make a teaching material for computer. That is to say, it will take a lot of time for an author to make a good teaching material including simulations so that the learner can master it easily. Because there is a large gap between the concepts of authors and the concepts for computer, it feels difficult to save more time for the author. In order to solve such a problem, we think the ontology technology is needed, and it is now widely accepted among the researchers. The system which we call SmartTrainer is a CBT (Computer Based Training) system including a set of simulators for facility of electric power systems. The development of SmartTrainer is partly motivated by the need for a powerful authoring system of Intelligent CBT systems. Of course, here we do not just want to discuss how to make an authoring system itself which can be used easily only, we also emphasize on what the simulation is from the view of training task. In this paper, as a preliminary step, we will mention the intelligent training system itself, and then explain what is ontology which is used as a base for making our Intelligent Training System. At last, we will briefly discuss the designing of authoring environment based on ontology in Smart Trainer. 2 Intelligent Training System General, Intelligent Training System is composed of student module, teaching materials module, teaching module and simulation module. Many AI-oriented researchers have the ambitious and visionary goal to supplant experienced human instructors by aiming for stand alone training systems(Steamer and Sherlock systems). The training system developed by LeiLa and Richard envisage serves instead as training support tools, tools that assist human instructors to perform their task more effectively[Leila, 1996]. In this paper, we also try to make an environment that assist the author make the simulation-based teaching material easily based on ontology. As the characteristics of our training system, Practice->knowledge->Practice is an important one. For the first practice, a learner can do simulation by himself/herself to solve a problem, if he/she made a mistake, system will teach him/her the principle behind the operating to overcome the error, and the last practice can check the learning effect of the learner. As a simulation-based training system, for the system, when to start a simulation; for the teaching material author, how to set the simulation conveniently are the subjects which we will discuss here. Because our authoring environment for Intelligent Training System is based on ontology, in next session, we will explain what ontology is? Who use or write ontology? and what the role of ontology for building an Intelligent Training System is.