부산대학교 도서관

Virtual Humans' Motivation Generation from Personality and Environment Wei Song, Kyungeun Cho, Kyhyun Um, Christian Darken Virtual agents are becoming more and more intelligent and adaptive to the dynamic virtual environment. The agents must make behavior planning based on characteristics that differ from one to the other. This paper describes a personality-based motivation oriented behavior planning system for an autonomous agent in games and virtual environments. The motivation selection process is implemented by evaluating and extracting motivation. The motivation extraction by both interception with personality variable and probability distribution allows the NPC to make multiple decisions in certain situations in order to embody realistic virtual humans. This model increases the complexity of the virtual human's action selection by adding personality attributes. With this decision making system, the virtual agent can specify motivation autonomously according to its characteristics. The proposed behavior planning system was tested in the restaurant simulation game "Crazy Waitress" to elucidate the mechanism of the framework.