Some Approaches to Rational Verification in Multiagent Systems
Date:
Recently, with the rapid advances of artificial intelligence, many researchers from verification community are starting to work on the analysis of systems composed of (semi)autonomous components known as multiagent systems. With this increasing interest, many concepts for reasoning about the behaviour of such systems are proposed. Among them is rational verification which is concerned with establishing whether a property can be sustained in a system composed of rational agents. In our research, we study different approaches to realise the notion of rational verification and strive for a concrete tool implementing the paradigm. We begin by introducing a formal framework as the foundation of our approaches. We then discuss the ability of current techniques/tools to perform rational verification and present some methods we have developed to expand the limits. We conclude with our ongoing work and possible future directions. Extended Abstract Slides