Search for an Appropriate Behavior within the Emotional Regulation in Virtual Creatures Using a Learning Classifier...

Area of Science:

• Artificial Intelligence
• Cognitive Science
• Neuroscience

Background:

• Emotion regulation involves conscious or unconscious control of emotional behaviors.
• Conscious emotion regulation relies on prior experiences and situational adaptation.
• Current literature lacks methods for endowing virtual agents with emotion regulation.

Purpose of the Study:

• To develop a computational model for emotion regulation in virtual creatures.
• To enable virtual agents to compute appropriate behaviors in specific emotional situations.
• To address the gap in artificial emotional intelligence research.

Main Methods:

• Utilized a learning classifier system (LCS) to model emotion regulation.
• Implemented a system capable of adapting behaviors based on learned experiences.
• Tested the model in both known and unknown situational contexts.

Main Results:

• The proposed LCS model successfully generated appropriate behaviors in emotional situations.
• The system demonstrated effective adaptation even in scenarios lacking prior experience.
• The approach proved powerful in situations with available prior information.

Conclusions:

• It is feasible to computationally model and implement emotion regulation in virtual agents.
• Learning classifier systems offer a viable mechanism for artificial emotional intelligence.
• The developed method advances the capability of virtual creatures to exhibit adaptive emotional behaviors.