Alexander Kraus und Han van der Aa have received the Best Student Paper Award at the 22nd Business Process Management Conference 2024 in Krakow, for their paper entitled “Looking for Change: A Computer Vision Approach for Concept Drift Detection in Process Mining”.
The paper presents a novel approach for detecting concept drifts that follows a different paradigm. Instead of using heuristic (unsupervised) techniques to identify drifts in event logs, the authors propose to use a machine learning model from the field of computer vision after transforming an event log into an image that encodes evolution of the recorded business process. The model is then trained on a large, labeled dataset, enabling it to learn directly how drifts manifest themselves in event logs. As a result, the computer vision-based approach achieves higher accuracy and greater robustness to noise compared to state-of-the-art solutions.
Interested? Click here for the full paper.