Avoiding over-fitting in ILP-based process discovery

S.J. Zelst, van, B.F. Dongen, van, W.M.P. van der Aalst

Research output: Chapter in Book/Report/Conference proceedingConference contributionAcademicpeer-review

34 Citations (Scopus)
1 Downloads (Pure)


The aim of process discovery is to discover a process model based on business process execution data, recorded in an event log. One of several existing process discovery techniques is the ILP-based process discovery algorithm. The algorithm is able to unravel complex process structures and provides formal guarantees w.r.t. the model discovered, e.g., the algorithm guarantees that a discovered model describes all behavior present in the event log. Unfortunately the algorithm is unable to cope with exceptional behavior present in event logs. As a result, the application of ILP-based process discovery techniques in everyday process discovery practice is limited. This paper addresses this problem by proposing a filtering technique tailored towards ILP-based process discovery. The technique helps to produce process models that are less over-fitting w.r.t. the event log, more understandable, and more adequate in capturing the dominant behavior present in the event log. The technique is implemented in the ProM framework. Keywords: Process mining Process discovery Integer linear programming Filtering
Original languageEnglish
Title of host publicationBusiness Process Management
Subtitle of host publication13th International Conference, BPM 2015, Innsbruck, Austria, August 31 - September 3, 2015, Proceedings
EditorsH.R. Motahari-Nezhad, J. Recker, M. Weidlich
Place of PublicationDordrecht
ISBN (Electronic)978-3-319-23063-4
ISBN (Print)978-3-319-23062-7
Publication statusPublished - 2015

Publication series

NameLecture Notes in Computer Science
ISSN (Print)0302-9743


Dive into the research topics of 'Avoiding over-fitting in ILP-based process discovery'. Together they form a unique fingerprint.

Cite this