Concept drifts detection and localisation in process mining

Abstract

Process mining provides methods and techniques for analyzing eventlogs recorded in modern information systems that support real-world operations. While analyzing an event-log, techniques in process mining assumes that the process as a static entity. This is not often the case due to possibility of phenomenon called concept drift. During the period of execution, process can experience concept drift and can evolve with respect to any of its associated perspectives exhibiting various patterns-of-change with different pace. This paper presents the method for detecting and localizing the sudden concept drifts in control-flow perspective of the process by using features extracted by processing the traces in process-log. © 2016 International Information Institute.