Dirk Fahland         Wil M.P. van der Aalst          SimplifyingMined Process Models
Process Mining, Currentlyevent   process mining   process log       algorithm      model                                  ...
Process Mining, Currently                                   readableevent   process mining   process    process log       ...
Post-Process the Model                                                          readable event     process mining    proce...
…Based on Original Event Log           process mining    process              algorithm       model                       ...
Analysis          process mining   process             algorithm      model                    readable event             ...
Idea: Re-Adjust Generalization             process mining       process                algorithm          model           ...
Unfold a Spaghetti-Model                           PAGE 7
Unfold Model wrt. a Log                                A                 ABDA                 ABCBDA                 ABCBC...
Unfold Model wrt. a Log                 unfold    A                                A                 ABDA                 ...
Unfold Model wrt. a Log                 unfold      A                                A                 ABDA               ...
Unfold Model wrt. a Log                 unfold        A                                A                 ABDA             ...
Unfold Model wrt. a Log                       unfold        A                                      A                      ...
Represents Concurrency                       unfold        A                                       A                      ...
Represents Concurrency        A                            AEBDA                            ABECBDA        B   E          ...
Represents Concurrency        A                       AEBDA                       ABECBDA        B   E          ABCBC     ...
Fold an unfolded model        A                merge equivalent nodes        B   E   necessary condition on               ...
Fold an unfolded model        A                merge equivalent nodes        B   E   necessary condition on               ...
Fold an unfolded model        A                merge equivalent nodes        B   E   necessary condition on               ...
Fold an unfolded model        A                merge equivalent nodes        B   E                     A    C   D         ...
Unfolding and Refolding                     unfold                                                A          fold         ...
Next: Simplifying and Generalizing                                                      readable                          ...
Implied Places        A                           implied place                           • does not restrict transitions ...
Special: Implied Places and Folding    A                     Ap                     p                                     ...
Configurable Simplification                                                               readable                        ...
ProM6 / Uma > www.processmining.org                                  PAGE 25
ProM6 / Uma > www.processmining.org                                  PAGE 26
ProM6 / Uma > www.processmining.org                                  PAGE 27
Experimental Results 15 benchmark logs, 6 industrial logs [www.promtools.org/prom5/]                                     ...
Experimental Results  15 benchmark logs, 6 industrial logs      [www.promtools.org/prom5/] model complexity = #arcs / #no...
Experimental Results precision: traces allowed by model and not in log 1.0 = only log behavior allowed rises/falls with...
from    tospaghetti   lasagna?
from    to less complexspaghetti   spaghetti
Lessons Learned techniques to navigate the model/behavior space use model and log together use model unfoldings break ...
And next?         process mining   process            algorithm      model               readable event                   ...
Dirk Fahland         about.me/dirk.fahland          SimplifyingMined Process Models
Upcoming SlideShare
Loading in …5
×

Simplifying Mined Process Models

775 views
685 views

Published on

This presentation was given by Dirk Fahland at the International Conference on Business Process Management 2011 (BPM'11) in Clermont-Ferrand, France on 31st August 2011.

Published in: Business, Technology
0 Comments
0 Likes
Statistics
Notes
  • Be the first to comment

  • Be the first to like this

No Downloads
Views
Total views
775
On SlideShare
0
From Embeds
0
Number of Embeds
16
Actions
Shares
0
Downloads
0
Comments
0
Likes
0
Embeds 0
No embeds

No notes for slide

Simplifying Mined Process Models

  1. 1. Dirk Fahland Wil M.P. van der Aalst SimplifyingMined Process Models
  2. 2. Process Mining, Currentlyevent process mining process log algorithm model PAGE 1
  3. 3. Process Mining, Currently readableevent process mining process process log algorithm model model PAGE 2
  4. 4. Post-Process the Model readable event process mining process simplify process log algorithm model modelcan replay the entire log introduce: post-processing can replay operations on the the entire log mined model PAGE 3
  5. 5. …Based on Original Event Log process mining process algorithm model readable event simplify process log modelcan replay the entire log introduce: post-processing can replay operations on the the entire log mined model PAGE 4
  6. 6. Analysis process mining process algorithm model readable event simplify process log model discover ordering relations  infer behavior behavior observed executions generalized behaviorincomplete knowledge PAGE 5
  7. 7. Idea: Re-Adjust Generalization process mining process algorithm model readable event simplify process log model unfold model wrt. log modelcomplexity fold, simplify, generalize behavior log PAGE 6
  8. 8. Unfold a Spaghetti-Model PAGE 7
  9. 9. Unfold Model wrt. a Log A ABDA ABCBDA ABCBC log C B D mined process model PAGE 8
  10. 10. Unfold Model wrt. a Log unfold A A ABDA ABCBDA B ABCBC log C B D A D mined process model PAGE 9
  11. 11. Unfold Model wrt. a Log unfold A A ABDA ABCBDA B ABCBC log C B C D B A D mined D process model A PAGE 10
  12. 12. Unfold Model wrt. a Log unfold A A ABDA ABCBDA B ABCBC log C B C D B A D mined D process modelC A PAGE 11
  13. 13. Unfold Model wrt. a Log unfold A A ABDA ABCBDA B ABCBC log C B C D B A D mined D B process modelCB A unfolding wrt. the log PAGE 12
  14. 14. Represents Concurrency unfold A A AEBDA ABECBDA B E ABCBC log C B E C D B A D mined D process modelC A unfolding wrt. the log PAGE 13
  15. 15. Represents Concurrency A AEBDA ABECBDA B E ABCBC log C D • is a process model B A • contains only behavior in the log • is acyclicC D • represents concurrency explicitly • labeled (several tasks with same label) A unfolding wrt. the log PAGE 14
  16. 16. Represents Concurrency A AEBDA ABECBDA B E ABCBC log C D unfold B A fold, simplify,C D generalize A unfolding wrt. the log PAGE 15
  17. 17. Fold an unfolded model A merge equivalent nodes B E necessary condition on equivalent transitions C D • same label B AC D A PAGE 16
  18. 18. Fold an unfolded model A merge equivalent nodes B E necessary condition on equivalent transitions C D • same label • equivalent pre-/post-places B AC D A PAGE 17
  19. 19. Fold an unfolded model A merge equivalent nodes B E necessary condition on equivalent transitions C D • same label • equivalent pre-/post-places B A various equivalences possible (see paper for some)C D A PAGE 18
  20. 20. Fold an unfolded model A merge equivalent nodes B E A C D C B E B A DC D A A PAGE 19
  21. 21. Unfolding and Refolding unfold A fold A C B E C B E D D refolded vs. original model • less behavior (replays the log and more) A • simpler structure PAGE 20
  22. 22. Next: Simplifying and Generalizing readable process simplify simplify process model model unfoldcomplexity fold simplify, generalize behavior log PAGE 21
  23. 23. Implied Places A implied place • does not restrict transitions B fold A remove from folded model C D • simpler model C B • same behavior B A D various techniques to findC D implied places A A PAGE 22
  24. 24. Special: Implied Places and Folding A Ap p A B C D C p fold D B C unfolding wrt. log folding may merge implied and non-implied places remove p: simpler model, more behavior (generalization) let user decide PAGE 23
  25. 25. Configurable Simplification readable process simplify simplify process model model unfoldcomplexity fold configurable simplify, generalize behavior log PAGE 24
  26. 26. ProM6 / Uma > www.processmining.org PAGE 25
  27. 27. ProM6 / Uma > www.processmining.org PAGE 26
  28. 28. ProM6 / Uma > www.processmining.org PAGE 27
  29. 29. Experimental Results 15 benchmark logs, 6 industrial logs [www.promtools.org/prom5/] PAGE 28
  30. 30. Experimental Results  15 benchmark logs, 6 industrial logs [www.promtools.org/prom5/] model complexity = #arcs / #nodes9.08.07.06.05.04.03.02.01.00.0 PAGE 29
  31. 31. Experimental Results precision: traces allowed by model and not in log 1.0 = only log behavior allowed rises/falls within limits (can be controlled) PAGE 30
  32. 32. from tospaghetti lasagna?
  33. 33. from to less complexspaghetti spaghetti
  34. 34. Lessons Learned techniques to navigate the model/behavior space use model and log together use model unfoldings break a rule and see what happens unfold modelcomplexity fold simplify, generalize behavior log PAGE 33
  35. 35. And next? process mining process algorithm model readable event simplify process log model process views most simple model covering 80% of the log improve mining algorithms? we showed: there is room for improvement PAGE 34
  36. 36. Dirk Fahland about.me/dirk.fahland SimplifyingMined Process Models

×