Andrea Burattin, Ph.D.

Assistant Professor

Online Conformance Checking for Petri Nets and Event Streams

Abstract

Within process mining, we can identify conformance checking as the task of computing the extent to which executions of a process model are in line with the reference behavior. Most approaches currently available in the literature (for imperative models, such as Petri nets) perform just a-posteriori analyses. This means that the amount of non-conformant behavior is quantified after the completion of the current execution. The tool presented in this paper, instead, proposes an approach for online conformance checking: not only it is capable of quantifying the deviating behavior on the fly, but the computation complexity is also restricted to a constant complexity per event analyzed. This enables the online analysis of an infinite stream of events. The tool is implemented as a package of the ProM framework and promising results have been obtained and are presented in this paper.

Paper Information and Files

Published in Online Proceedings of the BPM Demo Track 2017; Barcelona, Spain; September 10-15, 2017; CEUR-WS.org 2017.