Robust Drift Characterization from Event Streams of Business Processes

What is it about?

Process drifts are significant changes in the normal execution of a business process applied in response to changes in their operational environment, e.g., changes in workload, season, or regulations. Over time, these changes may negatively affect process performance. As such, several techniques have been proposed to detect process drifts. This paper takes the next step and contributes an efficient, accurate, and noise-tolerant automated method for explaining detected drifts with natural language statements.

Featured Image

Why is it important?

Process drifts may, over time, lead to process performance issues; hence, it is important to identify and understand them early on. Existing approaches for drift characterization are limited to simple changes that affect individual activities. The method proposed in this paper, which works both offline and online, is able to characterize complex drifts impacting process fragments of any size. Also, it relies on two cornerstone techniques, one to automatically discover process trees from event streams (logs) and the other to transform process trees using a minimum number of change operations. The latter presents a novel solution for the tree edit distance problem.