Title: A systematic survey of the business process mining-based approaches

Authors: Chaima Afifi; Ali Khebizi; Khaled Halimi

Addresses: Laboratory of Labstic, University 8 May 1945, Guelma, Algeria ' Department of Computer Science, University 8 May 1945, Guelma, Algeria ' Department of Computer Science, University 8 May 1945, Guelma, Algeria

Abstract: Nowadays, business processes (BPs) constitute the cornerstone of modern enterprise information systems. The behaviour conveyed by those BP expresses various endogenous factors reflecting companies' business rules and resources availability. Analysing these factors from a performance perspective of the BP execution data, is a challenging issue for both conventional process mining (PM) and PM-based deep learning (DL) methods. In this context, the multidisciplinary nature of the BP mining field, and the emergence of large panoply of BP mining methods, have induced a significant confusion in both academic and industrial communities. This paper addresses a thorough qualitative literature review of BP mining methods-based PM and DL approaches. Basing on a set of relevant properties, i.e., inputs, outputs, fundamental assumptions, algorithm type, and extra-functional properties, we have analysed the various studies having tackled the BP mining issue. Then, three assessment criteria have been established to evaluate the effectiveness of this article. At the end, the major limitations that are currently facing the BP mining field are identified and discussed.

Keywords: information systems; business process; event-log; process mining; machine learning; deep learning.

DOI: 10.1504/IJBPIM.2024.142658

International Journal of Business Process Integration and Management, 2024 Vol.11 No.4, pp.314 - 331

Received: 15 Mar 2024
Accepted: 22 Apr 2024
Published online: 14 Nov 2024 *

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