Towards an operational model for the propagation of public perception in multi-agent simulation