Finding process variants in event logs (short paper)
| Autor: | Bolt Iriondo, A.J., van der Aalst, W.M.P., de Leoni, M., Panetto, H., Debruyne, C., Gaaloul, W., Papazoglou, M., Paschke, A., Agostino Ardagna, C., Meersman, R. |
|---|---|
| Přispěvatelé: | Process Science |
| Jazyk: | angličtina |
| Rok vydání: | 2017 |
| Předmět: | |
| Zdroj: | On the Move to Meaningful Internet Systems. OTM 2017 Conferences: Confederated International Conferences: CoopIS, C&TC, and ODBASE 2017, Rhodes, Greece, October 23-27, 2017, Proceedings, Part I, 45-52 STARTPAGE=45;ENDPAGE=52;TITLE=On the Move to Meaningful Internet Systems. OTM 2017 Conferences |
| Popis: | The analysis of event data is particularly challenging when there is a lot of variability. Existing approaches can detect variants in very specific settings (e.g., changes of control-flow over time), or do not use statistical testing to decide whether a variant is relevant or not. In this paper, we introduce an unsupervised and generic technique to detect significant variants in event logs by applying existing, well-proven data mining techniques for recursive partitioning driven by conditional inference over event attributes. The approach has been fully implemented and is freely available as a ProM plugin. Finally, we validated our approach by applying it to a real-life event log obtained from a multinational Spanish telecommunications and broadband company, obtaining valuable insights directly from the event data. |
| Databáze: | OpenAIRE |
| Externí odkaz: |
|