Title: Machine learning models for predictive monitoring of business process execution delays

Authors: Walid Ben Fradj; Mohamed Turki; Faiez Gargouri

Addresses: Miracl Laboratory, ISIMS, University of Sfax, P.O. Box 242, 3021 Sfax, Tunisia ' Miracl Laboratory, ISIMS, University of Sfax, P.O. Box 242, 3021 Sfax, Tunisia ' Miracl Laboratory, ISIMS, University of Sfax, P.O. Box 242, 3021 Sfax, Tunisia

Abstract: Nowadays, organisations are increasingly aware of the importance of optimising the use of their knowledge resources and adopting a quality management model based on a process-centric approach. This approach requires a multidisciplinary approach that integrates the domains of knowledge management, business process management, and process mining. Thus, to enhance their performance and increase their responsiveness, organisations must identify, manage, and monitor all business processes (BPs) that may leverage crucial knowledge. It is imperative to implement a computerised system automating business processes to achieve these goals. In this context, we propose a new method for predicting the execution times of business processes, named BPETPM, based on the CRISP-DM approach. We employed machine learning techniques to exploit the execution data of a workflow engine. To demonstrate the relevance of this method, we developed an intelligent system for predicting BP execution times, called iBPMS4PET.

Keywords: business process management; BPM; process mining; knowledge management; KM; machine learning.

DOI: 10.1504/IJKMS.2026.153891

International Journal of Knowledge Management Studies, 2026 Vol.17 No.2, pp.207 - 226

Accepted: 09 Oct 2025
Published online: 29 May 2026 *

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