Extensive Survey on Methods and Techniques to Handle Concept Drift in Process Mining

Abstract

Business procedures benefit from process mining because it enables the extraction of crucial information from event logs to enhance process optimization. On-the-ground processes undergo evolution when stated as concept drift(cd) which makes the application of traditional process mining techniques become difficult. This paper presents a comprehensive investigation of the methods created to handle concept drift issues in process mining. Subsequently this research document will group these approaches according to their fundamental principles which include online learning approaches along with adaptive algorithms and hybrid methods. This paper explores the strengths and limitations of these methods when used to handle four main drift classifications: sudden, gradual, incremental, and recurring changes. The research survey presents upcoming tendencies and prospective future research routes while recognizing the need for strong time-sensitive scalable solutions to properly administer concept drift.