Process Mining Reloaded: Event Structures as a Unified Representation of Process Models and Event Logs
Download as PPT, PDF1 like2,387 views
The document discusses automated process discovery and conformance checking using Petri nets and event logs, focusing on methods such as the Alpha algorithm and various mining techniques. It highlights deviance mining and the application of discriminative sequence mining to capture process behavior and assess the fitness of models against event logs. Additionally, it explores log delta analysis for comparing business process event structures and the challenges in accurately modeling concurrency and causality.
![Deviance Mining
12
T1 <e11[d111:v111, …, d11n:v11n] e12[d121:v121, …, d12m:v12m] … e1p[d1p1:v1p1, …, d1pm:v1pm]>
…
Tq <eq1[dq11:vq11, …, dq1n:vq1n] eq2[dq21:vq21, …, dq2m:vq2m] … eqp[dqp1:vqp1, …, dqpm:vqpm]>
T1 <e11[d111:v111, …, d11n:v11n] e12[d121:v121, …, d12m:v12m] … e1p[d1p1:v1p1, …, d1pm:v1pm]>
…
Tq <eq1[dq11:vq11, …, dq1n:vq1n] eq2[dq21:vq21, …, dq2m:vq2m] … eqp[dqp1:vqp1, …, dqpm:vqpm]>
Find a function F: Trace Boolean (or probability [0…1])
s.t.
•F is an accurate approximation of the given labeling
•F is explainable, e.g. set of simple predicates](https://image.slidesharecdn.com/petrinets2015-dumas-150624094540-lva1-app6892/85/Process-Mining-Reloaded-Event-Structures-as-a-Unified-Representation-of-Process-Models-and-Event-Logs-12-320.jpg)
![Comparison of Event Structures
22
?
ES1
ES2
Armas-Cervantes et al. Behavioral Comparison of Process Models Based on […] Event Structures. BPM’2014
Partially
Synchronized
Product (PSP)](https://image.slidesharecdn.com/petrinets2015-dumas-150624094540-lva1-app6892/85/Process-Mining-Reloaded-Event-Structures-as-a-Unified-Representation-of-Process-Models-and-Event-Logs-20-320.jpg)
![Comparison of Event Structures
24
In ES1, tasks C and B are
mutually exclusive, while
in ES2, B precedes C
In ES1, tasks C and B are
mutually exclusive, while
in ES2, B precedes C
?
ES1
ES2
Armas-Cervantes et al. Behavioral Comparison of Process Models Based on […] Event Structures. BPM’2014](https://image.slidesharecdn.com/petrinets2015-dumas-150624094540-lva1-app6892/85/Process-Mining-Reloaded-Event-Structures-as-a-Unified-Representation-of-Process-Models-and-Event-Logs-22-320.jpg)
Process Mining Reloaded: Event Structures as a Unified Representation of Process Models and Event Logs
- 1. Marlon Dumas University of Tartu, Estonia Petri Nets 2015 | Brussels | 24 June 2015
- 2. Process Mining / event log discovered model Discovery Conformance Deviance Difference diagnostics Performanc e input model Enhanced model event log’ 2
- 3. Automated Process Discovery 3 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 - … … … …
- 4. Automated Process Discovery • Relations-based – Alpha 4
- 5. Alpha Algorithm • Direct successors: A > B, B > C, C > D, A > C, C > B, B > E, E > F C > E, E > G B > D A B C D A C B E F • Causality: A → B, C → D, A → C, B → E, C → E, E → F, E → G , B → D • Concurrency: B ║ C • Exclusiveness: all other pairs A B C E G A C B D 5
- 6. Alpha Relations Matrix A B C D E F G A # → → → # # # B ← # || → → # # C ← || # → → # # D # ← ← # # # # E # ← ← # # → → F # # # # ← # # G # # # # ← # # 6
- 7. A B C D E F G A # → → # # # # B ← # || → → # # C ← || # → → # # D # ← ← # # # # E # ← ← # # → → F # # # # ← # # G # # # # ← # # Alpha Algorithm – Patterns 7 ⇔ a→ b, a→ c, b ║ c
- 8. Automated Process Discovery • Relations-based – Alpha: lossy (Badouel, Petri Nets 2012) – Alpha++ – Heuristics miner (frequency information) • Genetic • Region theory • Petri net synthesis • Integer Linear Programming (ILP) • … 8
- 11. Alignment-Based Conformance Log Model A B C D EA B B C Alignment E Fitness Precision How much behavior of the log is captured by the model? How accurate is the model describing the log? Munoz-Gama et al. Petri nets 2013 11
- 12. Deviance Mining 12 T1 <e11[d111:v111, …, d11n:v11n] e12[d121:v121, …, d12m:v12m] … e1p[d1p1:v1p1, …, d1pm:v1pm]> … Tq <eq1[dq11:vq11, …, dq1n:vq1n] eq2[dq21:vq21, …, dq2m:vq2m] … eqp[dqp1:vqp1, …, dqpm:vqpm]> T1 <e11[d111:v111, …, d11n:v11n] e12[d121:v121, …, d12m:v12m] … e1p[d1p1:v1p1, …, d1pm:v1pm]> … Tq <eq1[dq11:vq11, …, dq1n:vq1n] eq2[dq21:vq21, …, dq2m:vq2m] … eqp[dqp1:vqp1, …, dqpm:vqpm]> Find a function F: Trace Boolean (or probability [0…1]) s.t. •F is an accurate approximation of the given labeling •F is explainable, e.g. set of simple predicates
- 13. Simple “timely” claims Simple “slow” claims Deviance Mining via Model Delta Analysis 13 Suriadi et al. Understanding Process Behaviours in a Large Insurance Company in Australia. CAiSE 2013
- 14. Deviance Mining via Model Delta Analysis 14
- 15. Deviance Mining via Sequence Classification • Apply discriminative sequence mining methods to extract features characteristic of one class • Build classification models (e.g. decision trees) • Extract difference diagnostics from classification model C. Sun et al. Mining explicit rules for software process evaluation. 15
- 16. No Unified Foundation ≠ 16
- 17. (Prime) Event Structures • Model of concurrency based on events (occurrences of actions) and three relations – Causality – Conflict – Concurrency 17
- 18. Petri Nets Event Structures 18 b a b c d d c b d d a b c d d c b d d 0 1 2 3 4 5 6 7 8 0 4 5 6 7 a b c d d d 9
- 19. Nets With Cycles Prefix Unfolding 21 Petri net NPetri net N Complete prefix unfolding Complete prefix unfolding Causality-preserving prefix unfolding Causality-preserving prefix unfolding
- 20. Comparison of Event Structures 22 ? ES1 ES2 Armas-Cervantes et al. Behavioral Comparison of Process Models Based on […] Event Structures. BPM’2014 Partially Synchronized Product (PSP)
- 21. PSP Difference Statements 23
- 22. Comparison of Event Structures 24 In ES1, tasks C and B are mutually exclusive, while in ES2, B precedes C In ES1, tasks C and B are mutually exclusive, while in ES2, B precedes C ? ES1 ES2 Armas-Cervantes et al. Behavioral Comparison of Process Models Based on […] Event Structures. BPM’2014
- 23. BP-Diff: BPMN model comparison 25http://diffbp-bpdiff.rhcloud.com/
- 24. Event Logs Event Structures B || C Concurrency Oracle Run Merger 55 22 33 26
- 25. Event Structures for Log Delta Analysis 27 van Beest et al. Log delta analysis: Interpretable differencing of business process event logs. BPM’2015
- 26. Event Structures for Log Delta Analysis In L1, task C can be skipped after B, whereas in L2 it cannot In L1, task C can be skipped after B, whereas in L2 it cannot van Beest et al. Log delta analysis: Interpretable differencing of business process event logs. BPM’2015 28
- 27. Log Delta Analysis vs. Sequence Classification 448 cases 7329 events 363 cases, 7496 events Sequence classification 106- 130 statements IF |“NursingProgressNotes”| > 7.5 THEN L1 IF |“Nursing Progress Notes”| ≤ 7.5 AND |“Nursing Assessment”| > 1.5 THEN L2 … Sequence classification 106- 130 statements IF |“NursingProgressNotes”| > 7.5 THEN L1 IF |“Nursing Progress Notes”| ≤ 7.5 AND |“Nursing Assessment”| > 1.5 THEN L2 … Log delta analysis 48 statements In L1, “Nursing Primary Assessment” is repeated after “Medical Assign Start” and “Triage Request”, while in L2 it is not. … Log delta analysis 48 statements In L1, “Nursing Primary Assessment” is repeated after “Medical Assign Start” and “Triage Request”, while in L2 it is not. … 29 van Beest et al. Log delta analysis: Interpretable differencing of business process event logs. BPM’2015
- 28. Event Structures for Conformance Checking 30 ABDE ADBE ACDE ADCE
- 29. Event Structures for Conformance Checking 31 In the model, task C and B are in conflict, whereas in the log, B precedes C In the model, task C and B are in conflict, whereas in the log, B precedes C
- 30. … vs. alignment-based conformance checking 32 ABDE ADBE ACDE ADCE ABCDE ABDCE ADBCE A B C D E A C D E A B D C E A B D E A D B C E A D C E ?
- 31. Event Structures for Process Discovery? 33 ABDE ACDE ACDF Fold
- 33. The Road Ahead • Developing more accurate concurrency oracles – Dealing with (short) loops in parallel branches • Defining folding operators to generalize & simplify Petri nets synthesized from ES – Controlled generalization • Extensions to events with data payloads 35
Editor's Notes
- #12: State-of-the-art conformance approach based on alignments Find the optimal trace in model that better describes each trace on the log. Occam RazorFitness Precision Other dimensions but here only this.