A Robust and Accurate Approach to Detect Process Drifts from Event Streams

TL;DR

This paper introduces a robust, accurate, and fast offline process drift detection method that effectively identifies behavioral changes in event streams, even in noisy data, using bidirectional search to locate both additions and removals of behaviors.

Contribution

The paper presents a novel offline process drift detection approach that accurately detects behavioral changes and is resilient to noise using bidirectional search techniques.

Findings

Accurately detects process drifts in event logs.

Robust to noisy data and maintains speed.

Validated on artificial and real-life logs.

Abstract

Business processes are bound to evolve as a form of adaption to changes, and such changes are referred as process drifts. Current process drift detection methods perform well on clean event log data, but the performance can be tremendously affected by noises. A good process drift detection method should be accurate, fast, and robust to noises. In this paper, we propose an offline process drift detection method which identifies each newly observed behaviour as a candidate drift point and checks if the new behaviour can signify significant changes to the original process behaviours. In addition, a bidirectional search method is proposed to accurately locate both the adding and removing of behaviours. The proposed method can accurately detect drift points from event logs and is robust to noises. Both artificial and real-life event logs are used to evaluate our method. Results show that our method can consistently report accurate process drift time while maintaining a reasonably fast detection speed.