Emotional Memory and Adaptive Personalities

摘要: Believable agents designed for long-term interaction with human users need to adapt them in a way which appears emotionally plausible while maintaining consistent personality. For short-term interactions restricted environments, scripting and state machine techniques can create emotion personality, but these methods are labor intensive, hard extend, brittle new environments. Fortunately, research memory, personality humans animals points solution this problem. Emotions focus an animal’s attention on things it needs care about, strong emotions trigger enhanced formation of enabling the animal its emotional response objects situations environment. In process becomes reflective: stress or frustration reevaluating past behavior respect personal standards, turn lead setting strategies goals. To aid authoring adaptive agents, we present artificial intelligence model inspired by psychological results triggers case-based preference learning behavioral adaptation guided models. Our tests robot pets embodied characters show that extend range increase sophistication agent without additional hand-crafted behaviors.