Continuous discovery of Causal nets for non-stationary business processes using the Online Miner

Jedrzej Potoniec, Daniel Sroka, Tomasz P. Pawlak

Abstract: Highlights•We propose Online Miner (OM) algorithm for the discovery of business process models.•Experiments involving 18 real-world event logs show quick model adaptation to process changes.•OM produces well-formed, perfect fit, and maximaly bound Causal nets.•The produced Causal nets come in handy to detect rare abuses, frauds, and defects.