Decision Mining Revisited - Discovering Overlapping Rules

Felix Mannhardt
Felix MannhardtResearch Scientist at SINTEF
Decision Mining Revisited
Discovering Overlapping Rules
Felix Mannhardt, Massimiliano de Leoni,
Hajo A. Reijers, Wil M.P. van der Aalst
Scope: Mining decision rules from event logs
PAGE 1
Apply
Amount
Grant
Extensive
Check
Reject
Eligibility
Simple
Check
Request
Information
Income
Receive
Information
Category
Activity
Data
Control-flow – Petri net defines order & possible choices
PAGE 2
Apply Grant
Extensive
Check
Reject
Simple
Check
Request
Information
Receive
Information
Exclusive
Choice
Sequence
Exclusive
Choice
Data-perspective – Data Petri Net modelling decisions
PAGE 3
Decision point
Data recording
Decision rule
PAGE 4
DMN 1.1 released on 2016
Widely adopted by tool vendors, for example:
U Eligibility Outcome
1 Yes Grant
2 No Reject
Decision Table
Grant
Reject
[Eligibility = No]
[Eligibility = Yes]
Comparing the Petri net notation to DMN
Decision Rule / Guard
Why are overlapping rules needed?
PAGE 5
Incomplete
Information
• Not recorded
• Process context
• Confidential
• ...
• Expert approval
• Deferred choice
• Randomized check
• Inconsistent human behavior
• ...
Goal: Discover rules which may overlap
PAGE 6
Process Model
Event Log
Process Model with
Overlapping Decision Rules
Overlapping Rule
Discovery
Decision point - Mutually-exclusive rule
PAGE 7
Grant
Reject
[Eligibility = No]
[Eligibility = Yes]
Count Eligibility Outcome
5x “No” Reject
20x “Yes” Grant
Observation instances from an event log
Grant
Reject
Decision point – Overlapping rule
PAGE 8
C Rating Amount Activity
1 Good - Simple Check
2 Bad - Extensive Check
3 Bad Low Simple Check
4 Bad High Request Information
5 Unknown - Request Information
Alternative Decision Table Notation
Proposed Discovery Method
PAGE 9
Process Model
Event Log
Process Model
With Overlapping Rules
Overlapping Rule
Discovery
foreach
Decision Point
Collect
Instances
1st
Classification
2nd
Classification
Collect
Misclassified
Build
Rules
1) Collect Instances
PAGE 10
Event Log collect
Rating Amount Outcome
6x Good Low Simple
6x Good High Simple
6x Bad High Extensive
4x Bad High Request
6x Bad Low Extensive
4x Bad Low Simple
6x Unknown High Request
Observation instances
• Cyclic Behavior
• Noise (Missing / Additional Events)
• Unassigned values
• Inconsistent recording
Alignment-based method
2) 1st Classification & 3) Misclassified Instances
PAGE 11
Rating Amount Outcome
6x Good Low Simple
6x Good High Simple
6x Bad High Extensive
4x Bad High Request
6x Bad Low Extensive
4x Bad Low Simple
6x Unknown High Request
Rating
Simple RequestExtensive
Good Unknown
Bad
12 OK 12 OK
8 NOK
6 OK
Instances Decision Tree
4) 2nd Classification
PAGE 12
Instances
Amount
Request Simple
High Low
2nd Decision Tree
Rating Amount Outcome
4x Bad High Request
4x Bad Low Simple
5) Build Overlapping Decision Rules
PAGE 13
Rating
Simple RequestExtensive
Good Unknown
Bad
Amount
Request Simple
High Low
Compiled to overlapping rules
If Rating = Good then Simple
If Rating = Unknown then Request
If Rating = Bad then Extensive
If Rating = Bad AND Amount = High
then Request
If Rating = Bad AND Amount = Low
then Simple
Resulting Data-aware Process Model
PAGE 14
Trade-off: Precise and fitting model
PAGE 15
Rating Amount Outcome
6x Good Low Simple
6x Good High Simple
6x Bad High Extensive
4x Bad High Request
6x Bad Low Extensive
4x Bad Low Simple
6x Unknown High Request
Unfitting
Imprecise
[Underfitting]
Good Trade-off
Evaluation – Measures
PAGE 16
Precision Fitness
How much unobserved
behavior is modelled?
How much observed
behavior is modelled?
Image source (CC BY-SA): https://en.wikipedia.org/wiki/Precision_and_recall#/media/File:Precisionrecall.svg
Evaluation – Setup
PAGE 17
Method Description Expected
Precision
Expected
Fitness
WO Without rules Poor Good
DTF Mutually-exclusive approach Good Poor
DTT Naïve overlapping approach Poor Good
DTO Presented overlapping approach Balanced Balanced
Dataset # Traces # Events # Attributes # Decisions
Road Fines 150,000 500,000 9 5
Hospital 1,000 15,000 39 11
Datasets
Compared Methods
Evaluation – Example rules in the hospital data
PAGE 18
Method Intensive Care Normal Care Skip
DTO L > 0 ∧ H = true L > 0 L ≤ 0 ∨
(L > 0 ∧ H = false)
DTT true L > 0 L ≤ 0
DTF false L > 0 L ≤ 0
Imprecise
Unfitting
Good
trade-off
Evaluation – Precision & Fitness
PAGE 19
Fitness Precision
• Fitness  how often rules are violated
• DTO improves fitness over DTF (mutually-exclusive)
• Precision  how strict are the rules
• DTO improves precision against WO
• DTO does sacrifice precision vs. DTF
Conclusion & Future Work
• Method: Discovery of overlapping rules using event logs
• Based on decision tree induction
• ProM framework: MultiPerspectiveExplorer
http://www.promtools.org
• Results: Trade-off fitness & precision
• Improves the model fitness over
standard trees
• Improves the model precision over
naïve approach
• Future work
• Better experimental validation
• Manage the complexity of discovered rules
• Imbalanced distributions
PAGE 20
Questions?
PAGE 21
@fmannhardt - f.mannhardt@tue.nl - http://promtools.org
Multi-Perspective Explorer
1 of 22

Recommended

Measuring the Precision of Multi-perspective Process Models by
Measuring the Precision of Multi-perspective Process ModelsMeasuring the Precision of Multi-perspective Process Models
Measuring the Precision of Multi-perspective Process ModelsFelix Mannhardt
930 views9 slides
From Low-Level Events to Activities - A Pattern based Approach by
From Low-Level Events to Activities - A Pattern based ApproachFrom Low-Level Events to Activities - A Pattern based Approach
From Low-Level Events to Activities - A Pattern based ApproachFelix Mannhardt
1.5K views23 slides
What is Data Science and How to Succeed in it by
What is Data Science and How to Succeed in itWhat is Data Science and How to Succeed in it
What is Data Science and How to Succeed in itKhosrow Hassibi
669 views50 slides
Causal reasoning and Learning Systems by
Causal reasoning and Learning SystemsCausal reasoning and Learning Systems
Causal reasoning and Learning SystemsTrieu Nguyen
1.1K views70 slides
Big Data Without Action Means Nothing by
Big Data Without Action Means NothingBig Data Without Action Means Nothing
Big Data Without Action Means NothingGlenn Pasch
535 views28 slides
Automated Decision making with Predictive Applications – Big Data Hamburg by
Automated Decision making with Predictive Applications – Big Data HamburgAutomated Decision making with Predictive Applications – Big Data Hamburg
Automated Decision making with Predictive Applications – Big Data HamburgLars Trieloff
1.6K views90 slides

More Related Content

Similar to Decision Mining Revisited - Discovering Overlapping Rules

Making ‘Big Data’ Your Ally – Using data analytics to improve compliance, due... by
Making ‘Big Data’ Your Ally – Using data analytics to improve compliance, due...Making ‘Big Data’ Your Ally – Using data analytics to improve compliance, due...
Making ‘Big Data’ Your Ally – Using data analytics to improve compliance, due...emermell
1K views40 slides
Automated decision making with predictive applications – Big Data Frankfurt by
Automated decision making with predictive applications – Big Data FrankfurtAutomated decision making with predictive applications – Big Data Frankfurt
Automated decision making with predictive applications – Big Data FrankfurtLars Trieloff
3.2K views82 slides
Ingenia consultants-9 basic steps towards TRM compliance by
Ingenia consultants-9 basic steps towards TRM complianceIngenia consultants-9 basic steps towards TRM compliance
Ingenia consultants-9 basic steps towards TRM complianceSami Benafia
92 views10 slides
Data analytics, a (short) tour by
Data analytics, a (short) tourData analytics, a (short) tour
Data analytics, a (short) tourVenkatesh Prasad Ranganath
1.6K views55 slides
Feature selection with imbalanced data in agriculture by
Feature selection with  imbalanced data in agricultureFeature selection with  imbalanced data in agriculture
Feature selection with imbalanced data in agricultureAboul Ella Hassanien
153 views21 slides
Building better products through Experimentation - SDForum Business Intellige... by
Building better products through Experimentation - SDForum Business Intellige...Building better products through Experimentation - SDForum Business Intellige...
Building better products through Experimentation - SDForum Business Intellige...Deepak Nadig
516 views23 slides

Similar to Decision Mining Revisited - Discovering Overlapping Rules(20)

Making ‘Big Data’ Your Ally – Using data analytics to improve compliance, due... by emermell
Making ‘Big Data’ Your Ally – Using data analytics to improve compliance, due...Making ‘Big Data’ Your Ally – Using data analytics to improve compliance, due...
Making ‘Big Data’ Your Ally – Using data analytics to improve compliance, due...
emermell1K views
Automated decision making with predictive applications – Big Data Frankfurt by Lars Trieloff
Automated decision making with predictive applications – Big Data FrankfurtAutomated decision making with predictive applications – Big Data Frankfurt
Automated decision making with predictive applications – Big Data Frankfurt
Lars Trieloff3.2K views
Ingenia consultants-9 basic steps towards TRM compliance by Sami Benafia
Ingenia consultants-9 basic steps towards TRM complianceIngenia consultants-9 basic steps towards TRM compliance
Ingenia consultants-9 basic steps towards TRM compliance
Sami Benafia92 views
Feature selection with imbalanced data in agriculture by Aboul Ella Hassanien
Feature selection with  imbalanced data in agricultureFeature selection with  imbalanced data in agriculture
Feature selection with imbalanced data in agriculture
Building better products through Experimentation - SDForum Business Intellige... by Deepak Nadig
Building better products through Experimentation - SDForum Business Intellige...Building better products through Experimentation - SDForum Business Intellige...
Building better products through Experimentation - SDForum Business Intellige...
Deepak Nadig516 views
Ipac 2014 by cwhms
Ipac 2014Ipac 2014
Ipac 2014
cwhms870 views
Improving Findability through Site Search Analytics by Louis Rosenfeld
Improving Findability through Site Search AnalyticsImproving Findability through Site Search Analytics
Improving Findability through Site Search Analytics
Louis Rosenfeld1.9K views
Big Data Analytics and Open Data by Sharjeel Imtiaz
Big Data Analytics and Open Data Big Data Analytics and Open Data
Big Data Analytics and Open Data
Sharjeel Imtiaz406 views
Anomaly detection Workshop slides by QuantUniversity
Anomaly detection Workshop slidesAnomaly detection Workshop slides
Anomaly detection Workshop slides
QuantUniversity5.6K views
Automated decision making using Predictive Applications – Big Data Paris by Lars Trieloff
Automated decision making using Predictive Applications – Big Data ParisAutomated decision making using Predictive Applications – Big Data Paris
Automated decision making using Predictive Applications – Big Data Paris
Lars Trieloff1.2K views
Diseconomies of Scale webinar by Locality
Diseconomies of Scale webinarDiseconomies of Scale webinar
Diseconomies of Scale webinar
Locality688 views
Tailoring machine learning practices to support prescriptive analytics by Adam Doyle
Tailoring machine learning practices to support prescriptive analyticsTailoring machine learning practices to support prescriptive analytics
Tailoring machine learning practices to support prescriptive analytics
Adam Doyle114 views
Optimizing Protocol Planning, Feasibility, and Site Selection through an Inte... by will buckley
Optimizing Protocol Planning, Feasibility, and Site Selection through an Inte...Optimizing Protocol Planning, Feasibility, and Site Selection through an Inte...
Optimizing Protocol Planning, Feasibility, and Site Selection through an Inte...
will buckley999 views
IT Operation Analytic for security- MiSSconf(sp1) by stelligence
IT Operation Analytic for security- MiSSconf(sp1)IT Operation Analytic for security- MiSSconf(sp1)
IT Operation Analytic for security- MiSSconf(sp1)
stelligence789 views
Process Mining Data-driven Process Improvement - idBigdata Meetup 17 Oct 2017 by Muhammad Faisal Reza
Process Mining Data-driven Process Improvement - idBigdata Meetup 17 Oct 2017Process Mining Data-driven Process Improvement - idBigdata Meetup 17 Oct 2017
Process Mining Data-driven Process Improvement - idBigdata Meetup 17 Oct 2017

More from Felix Mannhardt

A Taxonomy for Combining Activity Recognition and Process Discovery in Indust... by
A Taxonomy for Combining Activity Recognition and Process Discovery in Indust...A Taxonomy for Combining Activity Recognition and Process Discovery in Indust...
A Taxonomy for Combining Activity Recognition and Process Discovery in Indust...Felix Mannhardt
527 views20 slides
Estimating the Impact of Incidents on Process Delay - ICPM 2019 by
Estimating the Impact of Incidents on Process Delay - ICPM 2019Estimating the Impact of Incidents on Process Delay - ICPM 2019
Estimating the Impact of Incidents on Process Delay - ICPM 2019Felix Mannhardt
559 views24 slides
Data-driven Process Discovery - Revealing Conditional Infrequent Behavior fro... by
Data-driven Process Discovery - Revealing Conditional Infrequent Behavior fro...Data-driven Process Discovery - Revealing Conditional Infrequent Behavior fro...
Data-driven Process Discovery - Revealing Conditional Infrequent Behavior fro...Felix Mannhardt
1.1K views21 slides
Unsupervised Event Abstraction using Pattern Abstraction and Local Process Mo... by
Unsupervised Event Abstraction using Pattern Abstraction and Local Process Mo...Unsupervised Event Abstraction using Pattern Abstraction and Local Process Mo...
Unsupervised Event Abstraction using Pattern Abstraction and Local Process Mo...Felix Mannhardt
1.2K views11 slides
From Low-Level Events to Activities - A Pattern-based Approach by
From Low-Level Events to Activities - A Pattern-based ApproachFrom Low-Level Events to Activities - A Pattern-based Approach
From Low-Level Events to Activities - A Pattern-based ApproachFelix Mannhardt
143 views24 slides
Analyzing the Trajectories of Patients with Sepsis using Process Mining by
Analyzing the Trajectories of Patients with Sepsis using Process MiningAnalyzing the Trajectories of Patients with Sepsis using Process Mining
Analyzing the Trajectories of Patients with Sepsis using Process MiningFelix Mannhardt
1.2K views9 slides

More from Felix Mannhardt(7)

A Taxonomy for Combining Activity Recognition and Process Discovery in Indust... by Felix Mannhardt
A Taxonomy for Combining Activity Recognition and Process Discovery in Indust...A Taxonomy for Combining Activity Recognition and Process Discovery in Indust...
A Taxonomy for Combining Activity Recognition and Process Discovery in Indust...
Felix Mannhardt527 views
Estimating the Impact of Incidents on Process Delay - ICPM 2019 by Felix Mannhardt
Estimating the Impact of Incidents on Process Delay - ICPM 2019Estimating the Impact of Incidents on Process Delay - ICPM 2019
Estimating the Impact of Incidents on Process Delay - ICPM 2019
Felix Mannhardt559 views
Data-driven Process Discovery - Revealing Conditional Infrequent Behavior fro... by Felix Mannhardt
Data-driven Process Discovery - Revealing Conditional Infrequent Behavior fro...Data-driven Process Discovery - Revealing Conditional Infrequent Behavior fro...
Data-driven Process Discovery - Revealing Conditional Infrequent Behavior fro...
Felix Mannhardt1.1K views
Unsupervised Event Abstraction using Pattern Abstraction and Local Process Mo... by Felix Mannhardt
Unsupervised Event Abstraction using Pattern Abstraction and Local Process Mo...Unsupervised Event Abstraction using Pattern Abstraction and Local Process Mo...
Unsupervised Event Abstraction using Pattern Abstraction and Local Process Mo...
Felix Mannhardt1.2K views
From Low-Level Events to Activities - A Pattern-based Approach by Felix Mannhardt
From Low-Level Events to Activities - A Pattern-based ApproachFrom Low-Level Events to Activities - A Pattern-based Approach
From Low-Level Events to Activities - A Pattern-based Approach
Felix Mannhardt143 views
Analyzing the Trajectories of Patients with Sepsis using Process Mining by Felix Mannhardt
Analyzing the Trajectories of Patients with Sepsis using Process MiningAnalyzing the Trajectories of Patients with Sepsis using Process Mining
Analyzing the Trajectories of Patients with Sepsis using Process Mining
Felix Mannhardt1.2K views
XESLite - Handling Event Logs in ProM by Felix Mannhardt
XESLite - Handling Event Logs in ProMXESLite - Handling Event Logs in ProM
XESLite - Handling Event Logs in ProM
Felix Mannhardt1.4K views

Recently uploaded

ZEBRA FISH: as model organism.pptx by
ZEBRA FISH: as model organism.pptxZEBRA FISH: as model organism.pptx
ZEBRA FISH: as model organism.pptxmahimachoudhary0807
11 views17 slides
Vegetable grafting: A new crop improvement approach.pptx by
Vegetable grafting: A new crop improvement approach.pptxVegetable grafting: A new crop improvement approach.pptx
Vegetable grafting: A new crop improvement approach.pptxHimul Suthar
8 views69 slides
Evaluation and Standardization of the Marketed Polyherbal drug Patanjali Divy... by
Evaluation and Standardization of the Marketed Polyherbal drug Patanjali Divy...Evaluation and Standardization of the Marketed Polyherbal drug Patanjali Divy...
Evaluation and Standardization of the Marketed Polyherbal drug Patanjali Divy...Anmol Vishnu Gupta
7 views10 slides
IMMUNODIAGNOSTICS KITS.pdf by
IMMUNODIAGNOSTICS KITS.pdfIMMUNODIAGNOSTICS KITS.pdf
IMMUNODIAGNOSTICS KITS.pdfvetrivel303632
17 views10 slides
2. Natural Sciences and Technology Author Siyavula.pdf by
2. Natural Sciences and Technology Author Siyavula.pdf2. Natural Sciences and Technology Author Siyavula.pdf
2. Natural Sciences and Technology Author Siyavula.pdfssuser821efa
11 views232 slides
Effect of Integrated Nutrient Management on Growth and Yield of Solanaceous F... by
Effect of Integrated Nutrient Management on Growth and Yield of Solanaceous F...Effect of Integrated Nutrient Management on Growth and Yield of Solanaceous F...
Effect of Integrated Nutrient Management on Growth and Yield of Solanaceous F...SwagatBehera9
5 views36 slides

Recently uploaded(20)

Vegetable grafting: A new crop improvement approach.pptx by Himul Suthar
Vegetable grafting: A new crop improvement approach.pptxVegetable grafting: A new crop improvement approach.pptx
Vegetable grafting: A new crop improvement approach.pptx
Himul Suthar8 views
Evaluation and Standardization of the Marketed Polyherbal drug Patanjali Divy... by Anmol Vishnu Gupta
Evaluation and Standardization of the Marketed Polyherbal drug Patanjali Divy...Evaluation and Standardization of the Marketed Polyherbal drug Patanjali Divy...
Evaluation and Standardization of the Marketed Polyherbal drug Patanjali Divy...
2. Natural Sciences and Technology Author Siyavula.pdf by ssuser821efa
2. Natural Sciences and Technology Author Siyavula.pdf2. Natural Sciences and Technology Author Siyavula.pdf
2. Natural Sciences and Technology Author Siyavula.pdf
ssuser821efa11 views
Effect of Integrated Nutrient Management on Growth and Yield of Solanaceous F... by SwagatBehera9
Effect of Integrated Nutrient Management on Growth and Yield of Solanaceous F...Effect of Integrated Nutrient Management on Growth and Yield of Solanaceous F...
Effect of Integrated Nutrient Management on Growth and Yield of Solanaceous F...
SwagatBehera95 views
별헤는 사람들 2023년 12월호 전명원 교수 자료 by sciencepeople
별헤는 사람들 2023년 12월호 전명원 교수 자료별헤는 사람들 2023년 12월호 전명원 교수 자료
별헤는 사람들 2023년 12월호 전명원 교수 자료
sciencepeople68 views
Exploring the nature and synchronicity of early cluster formation in the Larg... by Sérgio Sacani
Exploring the nature and synchronicity of early cluster formation in the Larg...Exploring the nature and synchronicity of early cluster formation in the Larg...
Exploring the nature and synchronicity of early cluster formation in the Larg...
Sérgio Sacani1.4K views
A Ready-to-Analyze High-Plex Spatial Signature Development Workflow for Cance... by InsideScientific
A Ready-to-Analyze High-Plex Spatial Signature Development Workflow for Cance...A Ready-to-Analyze High-Plex Spatial Signature Development Workflow for Cance...
A Ready-to-Analyze High-Plex Spatial Signature Development Workflow for Cance...
InsideScientific115 views
selection of preformed arch wires during the alignment stage of preadjusted o... by MaherFouda1
selection of preformed arch wires during the alignment stage of preadjusted o...selection of preformed arch wires during the alignment stage of preadjusted o...
selection of preformed arch wires during the alignment stage of preadjusted o...
MaherFouda17 views
Best Hybrid Event Platform.pptx by Harriet Davis
Best Hybrid Event Platform.pptxBest Hybrid Event Platform.pptx
Best Hybrid Event Platform.pptx
Harriet Davis8 views
Study on Drug Drug Interaction Through Prescription Analysis of Type II Diabe... by Anmol Vishnu Gupta
Study on Drug Drug Interaction Through Prescription Analysis of Type II Diabe...Study on Drug Drug Interaction Through Prescription Analysis of Type II Diabe...
Study on Drug Drug Interaction Through Prescription Analysis of Type II Diabe...
Factors affecting fluorescence and phosphorescence.pptx by SamarthGiri1
Factors affecting fluorescence and phosphorescence.pptxFactors affecting fluorescence and phosphorescence.pptx
Factors affecting fluorescence and phosphorescence.pptx
SamarthGiri17 views
Small ruminant keepers’ knowledge, attitudes and practices towards peste des ... by ILRI
Small ruminant keepers’ knowledge, attitudes and practices towards peste des ...Small ruminant keepers’ knowledge, attitudes and practices towards peste des ...
Small ruminant keepers’ knowledge, attitudes and practices towards peste des ...
ILRI9 views

Decision Mining Revisited - Discovering Overlapping Rules

  • 1. Decision Mining Revisited Discovering Overlapping Rules Felix Mannhardt, Massimiliano de Leoni, Hajo A. Reijers, Wil M.P. van der Aalst
  • 2. Scope: Mining decision rules from event logs PAGE 1 Apply Amount Grant Extensive Check Reject Eligibility Simple Check Request Information Income Receive Information Category Activity Data
  • 3. Control-flow – Petri net defines order & possible choices PAGE 2 Apply Grant Extensive Check Reject Simple Check Request Information Receive Information Exclusive Choice Sequence Exclusive Choice
  • 4. Data-perspective – Data Petri Net modelling decisions PAGE 3 Decision point Data recording Decision rule
  • 5. PAGE 4 DMN 1.1 released on 2016 Widely adopted by tool vendors, for example: U Eligibility Outcome 1 Yes Grant 2 No Reject Decision Table Grant Reject [Eligibility = No] [Eligibility = Yes] Comparing the Petri net notation to DMN Decision Rule / Guard
  • 6. Why are overlapping rules needed? PAGE 5 Incomplete Information • Not recorded • Process context • Confidential • ... • Expert approval • Deferred choice • Randomized check • Inconsistent human behavior • ...
  • 7. Goal: Discover rules which may overlap PAGE 6 Process Model Event Log Process Model with Overlapping Decision Rules Overlapping Rule Discovery
  • 8. Decision point - Mutually-exclusive rule PAGE 7 Grant Reject [Eligibility = No] [Eligibility = Yes] Count Eligibility Outcome 5x “No” Reject 20x “Yes” Grant Observation instances from an event log Grant Reject
  • 9. Decision point – Overlapping rule PAGE 8 C Rating Amount Activity 1 Good - Simple Check 2 Bad - Extensive Check 3 Bad Low Simple Check 4 Bad High Request Information 5 Unknown - Request Information Alternative Decision Table Notation
  • 10. Proposed Discovery Method PAGE 9 Process Model Event Log Process Model With Overlapping Rules Overlapping Rule Discovery foreach Decision Point Collect Instances 1st Classification 2nd Classification Collect Misclassified Build Rules
  • 11. 1) Collect Instances PAGE 10 Event Log collect Rating Amount Outcome 6x Good Low Simple 6x Good High Simple 6x Bad High Extensive 4x Bad High Request 6x Bad Low Extensive 4x Bad Low Simple 6x Unknown High Request Observation instances • Cyclic Behavior • Noise (Missing / Additional Events) • Unassigned values • Inconsistent recording Alignment-based method
  • 12. 2) 1st Classification & 3) Misclassified Instances PAGE 11 Rating Amount Outcome 6x Good Low Simple 6x Good High Simple 6x Bad High Extensive 4x Bad High Request 6x Bad Low Extensive 4x Bad Low Simple 6x Unknown High Request Rating Simple RequestExtensive Good Unknown Bad 12 OK 12 OK 8 NOK 6 OK Instances Decision Tree
  • 13. 4) 2nd Classification PAGE 12 Instances Amount Request Simple High Low 2nd Decision Tree Rating Amount Outcome 4x Bad High Request 4x Bad Low Simple
  • 14. 5) Build Overlapping Decision Rules PAGE 13 Rating Simple RequestExtensive Good Unknown Bad Amount Request Simple High Low Compiled to overlapping rules If Rating = Good then Simple If Rating = Unknown then Request If Rating = Bad then Extensive If Rating = Bad AND Amount = High then Request If Rating = Bad AND Amount = Low then Simple
  • 16. Trade-off: Precise and fitting model PAGE 15 Rating Amount Outcome 6x Good Low Simple 6x Good High Simple 6x Bad High Extensive 4x Bad High Request 6x Bad Low Extensive 4x Bad Low Simple 6x Unknown High Request Unfitting Imprecise [Underfitting] Good Trade-off
  • 17. Evaluation – Measures PAGE 16 Precision Fitness How much unobserved behavior is modelled? How much observed behavior is modelled? Image source (CC BY-SA): https://en.wikipedia.org/wiki/Precision_and_recall#/media/File:Precisionrecall.svg
  • 18. Evaluation – Setup PAGE 17 Method Description Expected Precision Expected Fitness WO Without rules Poor Good DTF Mutually-exclusive approach Good Poor DTT Naïve overlapping approach Poor Good DTO Presented overlapping approach Balanced Balanced Dataset # Traces # Events # Attributes # Decisions Road Fines 150,000 500,000 9 5 Hospital 1,000 15,000 39 11 Datasets Compared Methods
  • 19. Evaluation – Example rules in the hospital data PAGE 18 Method Intensive Care Normal Care Skip DTO L > 0 ∧ H = true L > 0 L ≤ 0 ∨ (L > 0 ∧ H = false) DTT true L > 0 L ≤ 0 DTF false L > 0 L ≤ 0 Imprecise Unfitting Good trade-off
  • 20. Evaluation – Precision & Fitness PAGE 19 Fitness Precision • Fitness  how often rules are violated • DTO improves fitness over DTF (mutually-exclusive) • Precision  how strict are the rules • DTO improves precision against WO • DTO does sacrifice precision vs. DTF
  • 21. Conclusion & Future Work • Method: Discovery of overlapping rules using event logs • Based on decision tree induction • ProM framework: MultiPerspectiveExplorer http://www.promtools.org • Results: Trade-off fitness & precision • Improves the model fitness over standard trees • Improves the model precision over naïve approach • Future work • Better experimental validation • Manage the complexity of discovered rules • Imbalanced distributions PAGE 20
  • 22. Questions? PAGE 21 @fmannhardt - f.mannhardt@tue.nl - http://promtools.org Multi-Perspective Explorer

Editor's Notes

  1. I would like to present our work about “Decision Mining – Discovering Overlapping Rules”. My name is Felix Mannhardt, I’m a PhD student from the Eindhoven University of Technology. This is joint work with Massimiliano, Hajo and Wil.
  2. First to scope our work, I would like to introduce some of the assumptions/notations underlying our work. We want to analyze decisions that took place in processes. We assume that processes can be represented by process models. Notation: Activities boxes Data rounded boxes
  3. The control-flow of a process can be described with process A process model, such as a Petri net, defines the ordering and dependencies between activities We choose Petri net as notation to be independent from the actual process modelling language (such as BPMN or similar) For example: …
  4. Next to the order and dependencies between activities: decisions are at the heart of processes For example, data is recorded during the execution of activities; exclusive-choice in the process are decision points; decision rules govern which activities can be executed
  5. Decision point, exclusive choice between two activities Mutually-exclusive rule defined
  6. - DMN decision table using the Collect hit policy
  7. Public ‘Road Fines” dataset, IEEE taskforce Private hospital dataset
  8. Simplified Model of the care-path at the hospital DTO get better scores for fitness and precision compared to the DTT Lactate level are related to admission,