Learning Colored Petri Nets Using Object-Centric Event Data (OCED2CPN)

IEEE Congress on Information Science and Technology (CiSt), pp. 1–6

Abstract

Traditional process mining assumes that each event is related to precisely one case. This can be compared to mainstream modeling notations using Petri nets (in particular WF-nets), BPMN diagrams, directly-follows graphs, flowcharts, or UML activity diagrams. These notations describe the life-cycle of a process instance (often called case in process mining). Currently, we see an uptake of Object-Centric Process Mining (OCPM), where events can refer to any number of objects and where objects can be related. OCPM includes process discovery starting from Object-Centric Event Data (OCED) to produce process models that describe different object types in a single diagram. Since Colored Petri Nets (CPNs) have been around for decades and have been widely used for modeling, verification, and simulation, CPNs are an obvious target model for OCPM. The transition from traditional process mining to OCPM can be compared with the transition from classical Petri nets to CPNs. However, it turns out to be very difficult to discover arbitrary CPNs. This keynote paper summarizes what has been done before and why it is challenging to discover CPNs. However, starting from OCED, subclasses of CPNs can be discovered (which we refer to as OCED2CPN). These insights are also relevant for conformance checking and forward-looking forms of process mining starting from OCED.

Authors 1

1. Wil M. P. van der Aalst corresponding Aachen Process and Data Science (PADS)

   RWTH Aachen University

   Affiliation as printed

   RWTH Aachen University,Process and Data Science (PADS),Aachen,Germany

   Process and Data Science (PADS), RWTH Aachen University, Aachen, Germany