A Survey on Concept Drift in Process Mining

TL;DR

This survey reviews how concept drift affects process mining, highlighting the challenges, existing detection techniques, and the need for standardized evaluation protocols in evolving process environments.

Contribution

It provides a comprehensive taxonomy of concept drift detection and online process mining techniques, and identifies gaps such as offline focus and evaluation inconsistencies.

Findings

PM mainly focuses on offline analysis

Evaluation of drift detection is hindered by lack of standard datasets

Existing techniques are diverse but lack unified benchmarks

Abstract

Concept drift in process mining (PM) is a challenge as classical methods assume processes are in a steady-state, i.e., events share the same process version. We conducted a systematic literature review on the intersection of these areas, and thus, we review concept drift in process mining and bring forward a taxonomy of existing techniques for drift detection and online process mining for evolving environments. Existing works depict that (i) PM still primarily focuses on offline analysis, and (ii) the assessment of concept drift techniques in processes is cumbersome due to the lack of common evaluation protocol, datasets, and metrics.