Using Social Institutions to Guide Virtual Agent Behaviour

One requirement by which virtual environments (VEs) are judged, is the believability of the virtual agents (VAs). One aspect of believability, is that agent responses to situations should not create cognitive dissonance and thereby distract the observer. One approach to this problem is the use of institutional models, in conjunction with classical AI techniques, to achieve the appropriate recognition of complex situations and provide guidance on the subsequent choice of action(s). We present a distributed approach in which externally governed normaware agents are used to control the performance of the VE characters, creating the capacity to improve both the effectiveness of the decision-making and the believability of the characters. We aim to show that the combination of an institution providing social reasoning and BDI agents providing individual reasoning, establishes a framework for enhancing believability by the interplay between: (i) the institution and IVAs in VEs, and (ii) norms maintained by the institution and the mental states of IVAs. From an engineering point of view, the framework provides a separation of concerns because the BDI agent is augmented with the capacity to process social obligations, while the specification and verification of social structure resides in the (possibly several) institutional models. We illustrate our approach with two scenarios: one on queuing and another on inter-personal distance theory.