Social–Emotional Conversational Agents Based on Cognitive Architectures and Machine Learning

Large language models can recognize the topic of arbitrary utterances, as well as their emotional coloring, but do not understand the logic of emotions, even though they can often generate adequate responses in a given context. On the other hand, cognitive architectures such as eBICA are capable of modeling the dynamics of emotional states in the general case but require assistance in understanding the meaning of statements and generating responses to them. This paper presents a way to integrate large language models and eBICA, allowing them to complement each other. The “Virtual Receptionist” and “Virtual Psychodiagnostician” paradigms have been studied, and the results are encouraging.