Towards an emergent model of emotions

Abstract

Various attempts have been made throughout the history of artificial intelligence (AI) to incorporate emotional support in intelligent systems. Still some paths are left untraveled. This paper presents an attempt to model emotions as a complex system based on the hypothesis that mind state can be considered as an emergent phenomena resulting from autonomous, interactive, elementary entities called emotions. Inspiration for our approach is found in Buddhist philosophical concepts about thought formation by combination of emotions. Multi Agent Systems (MAS) technology has been used for the realization of above emotional model because interactions and emergence are intrinsic features of MAS. Capabilities of the model are exposed through a virtual agent application called EME who perceives, understands, uses and expresses emotions. Evaluation of the agent implementation has shown good results in achieving emotional intelligence with enhanced quality.