COBPAM - Behavior patterns mining and analysis for flexible processes

COBPAM is a plugin in the ProM framework [4] for mining and analyzing behavior patterns.

Event logs contain recorded data about business processes execution. Process Mining is the research discipline that analyzes such event logs and aims to discover models describing the unfolding of the process. Many algorithms were proposed but most of them don’t take into account cases of high irregularities between execution instances. We focus on such cases where the processes are unstructured and more exactly on a particular method to get insight from them. Namely, the mining of behavioral patterns. We propose a novel and more efficient algorithm that guarantees certain properties on the extracted patterns. We also propose a framework to analyze and retrieve such patterns in a contextual data-aware fashion manipulating correlation and causation. Lastly, we devise an advanced algorithm for the pattern discovery that is further optimized. It yields more concise and relevant results while offering a visualization interface for easy and interactive analysis. These algorithms were implemented as a plugin in the ProM framework [4] as the package BehavioralPatternMining.

References

[1] Mehdi Acheli, Behavioral Pattern Mining for Flexible Processes. (Fouille de Patterns Comportementaux dans le Contexte de Processus Flexibles). PSL University, Paris, France, 2021

[2] Mehdi Acheli, Daniela Grigori, Matthias Weidlich:
Discovering and Analyzing Contextual Behavioral Patterns From Event Logs. IEEE Trans. Knowl. Data Eng. 34(12): 5708-5721 (2022

[3] Mehdi Acheli, Daniela Grigori, Matthias Weidlich:
Efficient Discovery of Compact Maximal Behavioral Patterns from Event Logs. CAiSE 2019: 579-594

[4] Boudewijn F. van Dongen, Ana Karla A. de Medeiros, H. M. W. Verbeek, A. J. M. M. Weijters, Wil M. P. van der Aalst:
The ProM Framework: A New Era in Process Mining Tool Support. ICATPN 2005: 444-454