Creating an emotional space for artificial beings

Abstract

Interacting intelligently within a temporally dynamic environment calls for truly adaptive performances from artificial being. One mechanism currently being explored to produce an intuitive-like behaviour in A1 applications is that of emotion. Several popular psychological theories of emotion have been the basis for a number of artificial agent models that are capable of synthesizing emtions. However, whilst these models have proved successful, they have been built on theories that were not conceived for use in computer applications and thus lack or do not explain clearly the internal variables and mechanisms needed for adaptive applications. This paper examines another psychological theory built on discrete evaluations of cognitive appraisal dimensions that overcomes this problem and demonstrates an artificial agent that has had this theory integrated into its architecture.