Towards a theory of events

Event-driven systems are used in a wide range of applications such as responding to missile attacks, interdicting potential terrorists, exploiting arbitrage opportunities and responding to congestion in supply chains. The designs of event-driven systems vary widely because the costs and benefits to users of different applications are markedly different. This talk proposes a framework for unifying designs of different types of applications by representing the design problem as a constrained optimization and by defining interaction between components in distributed event-based systems in terms of a concept called "shared models." This talk is intended to suggest that there are concepts that unify analyses of a range of event-driven systems in nature, human social organizations and information technology systems.