Event-based modularization: how emergent behavioral patterns must be modularized?

Abstract

Nowadays, detecting emergent behavioral patterns in the environment, representing and manipulating them become the main focus of many software systems such as traffic monitoring systems, runtime verification techniques and self-adaptive systems. In this paper, we discuss the need for dedicated linguistic constructs to modularly represent emergent behavioral patterns and their lifetime semantics. We explain the shortcomings of current languages with this regard. Inspired from the evolution of procedural languages to object-oriented and aspect-oriented languages, we explain the concept of event-based modularization, which can be regarded as the successor of the aspect-oriented modularization for representing emergent behavioral patterns and their lifetime semantics. We report on our work on event modules and their successor gummy modules, which facilitate representing behavioral patterns as a holistic module that encapsulates its lifetime semantics.