Title: Comparative analysis of clustering-based remaining-time predictive process monitoring approaches

Authors: Niyi Ogunbiyi; Artie Basukoski; Thierry Chaussalet

Addresses: School of Computer Science and Engineering, University of Westminster, 309 Regent St., London W1B 2HW, UK ' School of Computer Science and Engineering, University of Westminster, 309 Regent St., London W1B 2HW, UK ' School of Computer Science and Engineering, University of Westminster, 309 Regent St., London W1B 2HW, UK

Abstract: Predictive process monitoring aims to accurately predict a variable of interest (e.g., remaining time) or the future state of the process instance (e.g., outcome or next step). Various studies have been explored to develop models with higher predictive power. However, comparing the various studies is difficult as different datasets, parameters and evaluation measures have been used. This paper seeks to address this problem with a focus on studies that adopt a clustering-based approach to predict the remaining time to the end of the process instance. A systematic literature review is undertaken to identify existing studies which adopt a clustering-based remaining-time predictive process monitoring approach and performs a comparative analysis to compare and benchmark the output of the identified studies using five real-life event logs.

Keywords: operational business process management; process monitoring; remaining-time predictive modelling.

DOI: 10.1504/IJBPIM.2021.124023

International Journal of Business Process Integration and Management, 2021 Vol.10 No.3/4, pp.230 - 241

Received: 21 Mar 2020
Accepted: 14 Sep 2020

Published online: 11 Jul 2022 *

Full-text access for editors Full-text access for subscribers Purchase this article Comment on this article