Agents, beliefs, and plausible behavior in a temporal setting
In this paper, we add a notion of plausible behavior to the branching-time logic CTL so that we obtain a language to reason about what can (or must) plausibly happen. Moreover, we propose a non-standard notion of beliefs, which is defined in terms of epistemic relations and plausibility and we investigate properties of plausibility, knowledge and beliefs in this new framework. In particular, we show that knowledge is still an S5 modality, and that beliefs satisfy axioms K45 in general, and KD45 for the class of so called plausibly serial models. Finally, we show that the relationship between knowledge and beliefs for plausibly serial models is very natural and reflects the initial intuition well. Keywords: multi-agent systems, temporal logic, plausibility, beliefs.
Nutzung und Vervielfältigung:
Alle Rechte vorbehalten