Weighing the Pros and Cons: Process Discovery with Negative Examples

Research output: Chapter in Book/Report/Conference proceedingArticle in proceedingsResearchpeer-review

Contemporary process discovery methods take as inputs only positive examples of process executions, and so they are one-class classification algorithms. However, we have found negative examples to also be available in industry, hence we propose to treat process discovery as a binary classification problem. This approach opens the door to many well-established methods and metrics from machine learning, in particular to improve the distinction between what should and should not be allowed by the output model. Concretely, we (1) present a formalisation of process discovery as a binary classification problem; (2) provide cases with negative examples from industry, including real-life logs; (3) propose the Rejection Miner binary classification procedure, applicable to any process notation that has a suitable syntactic composition operator; and (4) apply this miner to the real world logs obtained from our industry partner, showing increased output model quality in terms of accuracy and model size.

Original languageEnglish
Title of host publicationBusiness Process Management - 19th International Conference, BPM 2021, Proceedings
EditorsArtem Polyvyanyy, Moe Thandar Wynn, Amy Van Looy, Manfred Reichert
Publication date2021
ISBN (Print)9783030854683
Publication statusPublished - 2021
Event19th International Conference on Business Process Management, BPM 2021 - Rome, Italy
Duration: 6 Sep 202110 Sep 2021


Conference19th International Conference on Business Process Management, BPM 2021
SeriesLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume12875 LNCS

Bibliographical note

Publisher Copyright:
© 2021, Springer Nature Switzerland AG.

    Research areas

  • Binary classification, Labelled event logs, Negative examples, Process mining

ID: 282680828