Druckansicht der Internetadresse:

Faculty of Mathematics, Physics & Computer Science

Chair for Databases and Information Systems – Prof. Dr.-Ing. Stefan Jablonski

Print page



Paper accepted at ADBIS 2021


The paper "Cost-sensitive Predictive Business Process Monitoring" authored by Martin Käppel, Stefan Jablonski, and Stefan Schönig has been accepted for presentation and publication at the 25th European Conference on Advances in Databases and Information Systems (ADBIS 2021). Below you can find the abstract of the paper.

Cost-sensitive Predictive Business Process Monitoring

In predictive business process monitoring current and historical process data from event logs is used to predict the evolvement of running process instances. A wide number of machine learning approaches, especially different types of artificial neural networks, are successfully applied for this task. Nevertheless, experimental studies revealed that the resulting predictive models are not able to properly predict non-frequent activities. In this paper we investigate the usefulness of the concept of cost-sensitive learning, which introduces a cost model for different activities to better represent them in the training phase. An evaluation of this concept applied to common predictive monitoring approaches on various real life event logs shows encouraging results.

Facebook Twitter Youtube-Kanal Instagram Blog UBT-A Contact