Thema: Comprehensible Predictive Models for Business Processes
Martin Matzner is an assistant professor at the Department of Information Systems at the University of Münster, Germany. In 2012, he received the Ph.D. degree in Information Systems from the University of Münster for his work on the management of networked service business processes. His main research interests include business process management, business process analytics, and service management. In these areas, he concluded and currently manages a number of research projects funded by the European Union, the German Federal Government and industry. His work has been published in peer-reviewed academic journals and presented at major international conferences.
Patrick Delfmann is an associate professor at the Department of Information Systems at the University of Münster, Germany. Currently, he is on leave and works as an acting professor at the Department for IS Research at the University of Koblenz-Landau, Germany. He was the coordinator of several research projects, funded by national and international funding organizations. Patrick teaches at the Universities of Koblenz-Landau and Münster and he was a visiting professor in Moscow (RU), Vienna (A), Biel (CH), and Osnabrück (D). His work comprises more than 90 scientific papers, many of which appeared in international journals (e.g., Information Systems, ISF, BISE, CAIS) and conference proceedings (e.g., ICIS, ECIS, ER). His research interests include conceptual modeling, model analysis, semantic process modeling, business process compliance management, business process improvement, and process mining.
Predictive modeling approaches in business process management provide a way to streamline operational business processes. For instance, they can warn decision-makers about undesirable events that are likely to happen in the future, giving the decision-makers an opportunity to intervene. The topic is gaining momentum in process mining, a field of research that has traditionally developed tools to discover business process models from datasets of past process behavior. Predictive modeling techniques are built on top of process-discovery algorithms. As these algorithms describe business process behavior using models of formal languages (e.g., Petri nets), strong language biases are necessary in order to generate models with the limited amounts of data included in the dataset. Naturally, corresponding predictive modeling techniques reflect these biases. Based on theory from grammatical inference, a field of research that is concerned with inducing language models, we design a new predictive modeling technique based on weaker biases. Fitting a probabilistic model to a dataset of past behavior makes it possible to predict how currently running process instances will behave in the future. To clarify how this technique works and to facilitate its adoption, we also design a way to visualize the probabilistic models. We assess the technique’s effectiveness in an experimental evaluation with synthetic and real-world data.