Complex Event Processing (CEP) enables analyses with high velocity on high volume and varied data. It is an established technology to enable flexible event-driven systems. These systems can provide real-time analytics for business process management or support the realization of cyber-physical systems in the context of the Internet of Things. While the technology is maturing rapidly, the design of event-based systems is still in its infancy and complex. In particular, there is no established means to sup-port the early stages of the IS development process through conceptual modeling. Currently, any com-prehensive graphical specification of the involved event processing networks (EPN) is tool-dependent and creates vendor lock-in effects before an informed decision for a software product can be made. Current conceptual modeling options are limited to generic flow diagrams and methods from related areas. In this paper, we describe a method to conceptually model EPN which allow not only the filter-ing and projection but also the translation, enrichment, and aggregation as well as the splitting and composition of events. This artifact enables the design of EPN prior to deciding for a particular CEP product alleviating vendor lock-in effects. We present meta models excerpts and a notation as well as sample code in Esper to emphasize their serializability. The results are applied to a sample process.
Janiesch, Christian and Diebold, Jakob, "CONCEPTUAL MODELING OF EVENT PROCESSING NETWORKS" (2016). Research Papers. 87.