Making Sense of Actor Behaviour

Sense-making with respect to actor-based systems is challenging because of the non-determinism arising from concurrent behaviour. One strategy is to produce a trace of event histories that can be processed post-execution. Given a semantic domain, the histories can be translated into visual representations of the semantics in the form of filmstrips. This paper proposes a general pattern for the production of filmstrips from actor histories that can be implemented in a way that is independent of the particular data types used to represent the events, semantics and graphical displays. We demonstrate the pattern with respect to a simulation involving predators and prey which is a typical agent-based application.