Construction of virtual KANSEI by Petri-net with GA and method of constructing personality

This study proposes a new method of constructing Virtual KANSEI (VK) in a narrow sense (VKINS) using the Petri-net with learning capability that is realized by the Genetic Algorithm (GA). The VKINS imitates human's KANSEI states (emotions, feelings and mental symbols). The Petri-net enables to visualize internal transition of emotions within the VKINS. The feasibility of this technique is exhibited by experiments with observed trajectories of human emotion. Moreover, the different personalities using the proposed VKINS are generated by tactical manipulations of transition firing rule in the Petri-net.