Events Put into Context (EPiC)

International Conference on Process Mining (ICPM), vol. 14, pp. 65–72

Abstract

Business process models can be (re)constructed using event data recorded during the process' execution. Similarly, event data can be used to verify conformance to prescribed behavior and to analyze and improve the underlying processes. However, not all events that are related to a process necessarily relate to its control-flow. Some events occur in the context of the process. In this work, we introduce the concept of context events to deal with these types of events. We show how distinguishing between contextual and control-flow events aids process discovery to obtain less complex process models. We demonstrate how visualizing context events on top of process models helps identify points in the process where context events occur often, aiding understanding. We analyze these benefits using two case studies involving real-life processes and event data.

Authors 3

1. Marcus Dees

   Eindhoven University of Technology · Uitvoeringsinstituut Werknemersverzekeringen

   Affiliation as printed

   Uitvoeringsinstituut Werknemersverzekeringen (UWV),Amsterdam,The Netherlands

   Eindhoven University of Technology, Eindhoven, The Netherlands

   Uitvoeringsinstituut Werknemersverzekeringen (UWV), Amsterdam, The Netherlands

2. Bart F. A. Hompes

   Eindhoven University of Technology

   Affiliation as printed

   Artifex Consultancy,Eindhoven,The Netherlands

   Artifex Consultancy, Eindhoven, The Netherlands

   Eindhoven University of Technology, Eindhoven, The Netherlands

3. Wil M. P. van der Aalst Aachen

   RWTH Aachen University · Eindhoven University of Technology · Fraunhofer Institute for Applied Information Technology

   Affiliation as printed

   RWTH Aachen University,Aachen,Germany

   Eindhoven University of Technology, Eindhoven, The Netherlands

   Fraunhofer Institute for Applied Information Technology

   RWTH Aachen University, Aachen, Germany