Beyond Tasks and Gateways: Automated Discovery of BPMN Models with Subprocesses, Boundary Events and Activity Markers
Download as PPT, PDF1 like2,081 views
The document explores automated process discovery in BPMN models, particularly focusing on subprocesses, boundary events, and activity markers. It discusses techniques for filtering irrelevant tasks, clustering traces, and extracting process hierarchies using various algorithms. The findings include evaluations of different models and algorithms, along with future directions for tool implementation and further developments in event-based functionality.
Beyond Tasks and Gateways: Automated Discovery of BPMN Models with Subprocesses, Boundary Events and Activity Markers
- 1. Beyond Tasks and Gateways: Discovering BPMN Models with subprocesses, boundary events and activity markers Raffaele Conforti, Marcello La Rosa Queensland University of Technology Marlon Dumas, Luciano García-Bañuelos University of Tartu BPM’2014 Conference, Eindhoven 11 September 2014 1
- 2. 2 Automated Process Discovery Enter Loan Application Retrieve Applicant Data Compute Installments Approve Simple Application Approve Complex Application Notify Rejection Notify Eligibility CID Task Time Stamp … 13219 Enter Loan Application 2007-11-09 T 11:20:10 - 13219 Retrieve Applicant Data 2007-11-09 T 11:22:15 - 13220 Enter Loan Application 2007-11-09 T 11:22:40 - 13219 Compute Installments 2007-11-09 T 11:22:45 - 13219 Notify Eligibility 2007-11-09 T 11:23:00 - 13219 Approve Simple Application 2007-11-09 T 11:24:30 - 13220 Compute Installements 2007-11-09 T 11:24:35 - … … … …
- 5. Automated Process Discovery: Handling Complexity Filter • Filter out “irrelevant” events (tasks) • Filter out “irrelevant” traces Abstract • Zoom into most frequent tasks or paths • Extract subprocesses Divide • Divide log by variants based on similarity (trace clustering) • Discover multiple process models rather than one 5
- 6. Related Work: ProM two-phase miner Bose, Veerbeck & van det Aalst: Discovering Hierarchical Process Models using ProM
- 7. ProM Two-Phase Miner 7 Produces this… Instead of…
- 8. Before 8
- 9. After 9
- 11. Data! 11
- 12. Extracting the Process Hierarchy Extract event tables ReceivePO Find primary keys Find foreign keys Cluster event types Split log per cluster 12 flat log log hierarchy Time POID Att2 Att3 11-24 17:12 1 … … 11-25 08:53 2 … … … … … … CreateMO Time MOID POID Att3 11-24 17:13 1 1 … 11-26 09:31 2 2 … … 2 … … ShipPO Time POID Att2 11-25 12:11 1 … … 2 … … … …
- 13. … the rest Discover one model per (sub- )process Identify interrupting boundary events Identify interrupting timer events Identify event subprocesses Identify loop/multi-instance markers 13 - Heuristics miner - ILP - Inductive Miner - Fodina Heuristics
- 14. Evaluation Setup Logs Traces Events Event types Events/type IWT (FRIS) 121 1472 13 113 Insurance 896 12437 9 1382 Order-to-cash 3000 32896 13 2530 • Four flat process discovery algos • Heuristics, ILP, Inductive (H), Fodina • Each algo with and without BPMN-Miner • Quality measures • Accuracy: Fitness, precision, F-score • Understandability: Size, Control-Flow Complexity, … 14
- 15. Evaluation - Results 15
- 16. Side-Effect: Correct Models 16
- 17. What’s Next • Standalone tool implementation • Currently in ProM nightly build • Further evaluation • Logs with larger number of event types • Noise resilience • Missing events can trick foreign key discovery • Further enrichment • Event-based gateways, more BPMN events… • Adding data conditions, completion conditions, … 17