Emotional biologically inspired cognitive architecture

Abstract

Abstract Human-like artificial emotional intelligence is vital for integration of future robots into the human society. This work introduces a general framework for representation and processing of emotional contents in a cognitive architecture, called “emotional biologically inspired cognitive architecture” (eBICA). Unlike in previous attempts, in this framework emotional elements are added virtually to all cognitive representations and processes by modifying the main building blocks of the prototype architectures. The key elements are appraisals associated as attributes with schemas and mental states, moral schemas that control patterns of appraisals and represent social emotions, and semantic spaces that give values to appraisals. Proposed principles are tested in an experiment involving human subjects and virtual agents, based on a simple paradigm in imaginary virtual world. It is shown that with moral schemas, but probably not without them, eBICA can account for human behavior in the selected paradigm. The model sheds light on clustering of social emotions and allows for their elegant mathematical description. The new framework will be suitable for implementation of believable emotional intelligence in artifacts, necessary for emotionally informed behavior, collaboration of virtual partners with humans, and self-regulated learning of virtual agents.