Towards Scalable Computational Models of Emotions for Autonomous Agents

Computational models of emotions (CMEs) are software systems designed to synthesize the mechanisms of the human emotion process. They are included in cognitive agent architectures to endow Autonomous Agents (AAs) with abilities for the evaluation of emotional stimuli, the simulation and expression of emotional feelings, and the development of emotionally driven responses. Although the literature reports several developments of CMEs, there is still a wide range of challenges that remain unaddressed regarding their development process. A key challenge is the development of scalable CMEs whose architecture is capable of implementing novel findings about human emotions. In this paper, we discuss the challenge of scalable CMEs and present a case study that demonstrates how the step by step application of a methodology that takes advantage of psychological and biological findings leads to the design of scalable CMEs. The results of this paper aim at promoting the development of AAs capable of meeting the complex requirements of current applications.