Weighing the Pros and Cons: Process Discovery with Negative Examples

Publikation: Bidrag til bog/antologi/rapportKonferencebidrag i proceedingsForskningfagfællebedømt

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.

OriginalsprogEngelsk
TitelBusiness Process Management - 19th International Conference, BPM 2021, Proceedings
RedaktørerArtem Polyvyanyy, Moe Thandar Wynn, Amy Van Looy, Manfred Reichert
ForlagSpringer
Publikationsdato2021
Sider47-64
ISBN (Trykt)9783030854683
DOI
StatusUdgivet - 2021
Begivenhed19th International Conference on Business Process Management, BPM 2021 - Rome, Italien
Varighed: 6 sep. 202110 sep. 2021

Konference

Konference19th International Conference on Business Process Management, BPM 2021
LandItalien
ByRome
Periode06/09/202110/09/2021
NavnLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Vol/bind12875 LNCS
ISSN0302-9743

Bibliografisk note

Publisher Copyright:
© 2021, Springer Nature Switzerland AG.

ID: 282680828