A General Framework for Correlating Business Process Characteristics

A General Framework for Correlating Business Process Characteristics Massimiliano de Leoni, Wil M. P. van der Aalst, Marcus Dees

Introduction Performance Related • What do the cases that are late have in Discovery common? • Do people work faster if they have more Informationwork? event log Conformance System Resource Related A B • What characterizes the workers that skip such-an-such activity? modelchecking. Correlation • Which types of claims are more prone to Analysis to wrong treatments by resources? Processlead Mining Cost and Risk Related • Which types of claims are more risky or more expensive? … A B model /

The problem of correlating process characteristic Relating Dependent any process characteristic to le b Independent ra u g fi n other process characteristics Co associated with vent Filterbinleg E ura g i f n given events Co le b a r u g nfi o C

A General framework for correlating process characteristics Dependent Characteristic, Independent Characteristics, Filtering Trace Manipulation, Manipulation Context Data If analysis needs to be refined Event Log 3. 2. Make Analysis Result 2. Manipulate and Enrich Event log • • • Conformance Resource Time Data Weather. . . Additional Objects Analysis Use Case Process Model . . . 1. Define Analysis Use Case Event Log

Example: Relating the cost of First Hospital Admission to the performing resource / 1 Case ID Timestamp 1 30 -11 -2011: 08. 27 1 2 -12 -2011: 13. 24 1 4 -12 -2011: 8. 30 1 4 -12 -2011: 13. 30 2 1 -12 -2011: 11. 00 2 2 -12 -2011: 15. 28 2 2 -12 -2011: 16. 35 2 3 -12 -2011: 13. 00 2 3 -12 -2011: 15. 00 2 4 -12 -2011: 9. 00 3 7 -12 -2011: 10. 00 3 8 -12 -2011: 13. 24 3 9 -12 -2011: 16. 35 4 6 -12 -2011: 14. 00 4 8 -12 -2011: 13. 24 4 10 -12 -2011: 16. 35 4 13 -12 -2011: 11. 00 4 13 -12 -2011: 16. 00 Activity First Hospital Admission Preoperative Screening Laparoscopic Gastrectomy Nursing First Hospital Admission Laparoscopic Gastrectomy Nursing First Hospital Admission Preoperative Screening Laparoscopic Gastrectomy First Hospital Admission Resource Cost Carol 90 Susanne 350 Andrew 500 Paul 250 Giuseppe 350 Simon 500 Clare 250 Paul 500 Andrew 250 Victor 34 Jane 200 Giulia 500 Paul 250 Gianluca 90 Robert 350 Giuseppe 350 Simon 500 Jane 300 • Dependent characteristic: Cost • Independent characteristics: Resource • Filtering: Retain all events for First Hospital Admission

Example: Relating the cost of First Hospital Admission to the performing resource / 2 Problem: Dependent characteristics defined over continuous domain Solution: Discretization techniques: equal-width or equal-frequency binning Resource!=“Jane” (50, 200) Resource=“Jane” (200, 300) • Dependent characteristic: Cost • Independent characteristics: Resource • Filtering: Retain all events for First Hospital Admission

Example: Decision Point Analysis What determines what to do as next when reaching a decision point? age salary amount length installment Decision Points

Example: Decision Point Analysis / 2 Trace augmentation is needed Case ID Timestamp 1 30 -11 -2011: 08. 27 1 2 -12 -2011: 13. 24 1 4 -12 -2011: 8. 30 1 4 -12 -2011: 13. 30 2 1 -12 -2011: 11. 00 2 2 -12 -2011: 15. 28 2 2 -12 -2011: 16. 35 2 3 -12 -2011: 13. 00 2 3 -12 -2011: 15. 00 2 4 -12 -2011: 9. 00 3 7 -12 -2011: 10. 00 3 8 -12 -2011: 13. 24 3 9 -12 -2011: 16. 35 4 6 -12 -2011: 14. 00 4 8 -12 -2011: 13. 24 4 10 -12 -2011: 16. 35 4 13 -12 -2011: 11. 00 4 13 -12 -2011: 16. 00 Activity First Hospital Admission Preoperative Screening Laparoscopic Gastrectomy Nursing First Hospital Admission Laparoscopic Gastrectomy Nursing First Hospital Admission Preoperative Screening Laparoscopic Gastrectomy First Hospital Admission Resource Cost Next Activity in Trace Carol 90 Preoperative Screening Susanne 350 Laparoscopic Gastrectomy Andrew 500 Nursing Paul 250 null Giuseppe 350 Laparoscopic Gastrectomy Simon 500 Nursing Clare 250 Laparoscopic Gastrectomy Paul 500 Nursing Andrew 250 First Hospital Admission Victor 34 null Jane 200 Laparoscopic Gastrectomy Giulia 500 Nursing Paul 250 null Gianluca 90 Preoperative Screening Robert 350 Preoperative Screening Giuseppe 350 Laparoscopic Gastrectomy Simon 500 First Hospital Admission Jane 300 null

Example: Decision Point Analysis / 2 Independent Characteristics Case ID Timestamp 1 30 -11 -2011: 08. 27 1 2 -12 -2011: 13. 24 1 4 -12 -2011: 8. 30 1 4 -12 -2011: 13. 30 2 1 -12 -2011: 11. 00 2 2 -12 -2011: 15. 28 2 2 -12 -2011: 16. 35 2 3 -12 -2011: 13. 00 2 3 -12 -2011: 15. 00 2 4 -12 -2011: 9. 00 3 7 -12 -2011: 10. 00 3 8 -12 -2011: 13. 24 3 9 -12 -2011: 16. 35 4 6 -12 -2011: 14. 00 4 8 -12 -2011: 13. 24 4 10 -12 -2011: 16. 35 4 13 -12 -2011: 11. 00 4 13 -12 -2011: 16. 00 Activity First Hospital Admission Preoperative Screening Laparoscopic Gastrectomy Nursing First Hospital Admission Laparoscopic Gastrectomy Nursing First Hospital Admission Preoperative Screening Laparoscopic Gastrectomy First Hospital Admission Dependent Characterist Resource Cost Next Activity in Trace Carol 90 Preoperative Screening Susanne 350 Laparoscopic Gastrectomy Andrew 500 Nursing Paul 250 null Giuseppe 350 Laparoscopic Gastrectomy Simon 500 Nursing Clare 250 Laparoscopic Gastrectomy Paul 500 Nursing Andrew 250 First Hospital Admission Victor 34 null Jane 200 Laparoscopic Gastrectomy Giulia 500 Nursing Paul 250 null Gianluca 90 Preoperative Screening Robert 350 Preoperative Screening Giuseppe 350 Laparoscopic Gastrectomy Simon 500 First Hospital Admission Jane 300 null 22 -2 -2021 PAGE 8

Example: Example Duration of case executions Dependent Characteristic Many well-studied problems are just ad-hoc instances for our framework

Implementation of the framework • Reference Implementation available for Pro. M 6. 4 • Check out the Feature. Prediction package of Pro. M • The main input is an event log • Additional input objects may be required for specific log manipulations Join the demo session on Wednesday from 16. 00 to 17. 30!!

A step-through guide: existing the duration of activities • Health-care Process enacted in a Dutch hospital. • How is the duration of Doctor appointments related to the patient characteristics and his/her treatment?

Selection of the characteristics to consider & Augmentation

Filtering on activity names We should only consider activity Afspraak (Dutch term for Appointment)

Choosing the dependent characteristic & parameter for constructing decision trees The dependent characteristic Parameters for building decision trees

Discretization n = Number of bins Equal-width binning The entire range is divided in n intervals of equal size Equal-frequency The entire range is divided in n intervals such that each interval contains the same number of observed values.

The resulting Decision Tree • Correlating the duration of execution of activity Afspraak to the other characterists of the process

The UWV case study • UWV is the institute that manage the unemployment benefits in the Netherlands. • UWV opens a reclamation when one receives unemployment benefits although not entitled. • UWV eventually discovers it. • In many cases, the excessive amount is not recollected. • UWV wants to predict the risk of eventual reclamations • For prediction, one needs to relate reclamations to the characterists of the customer and the process execution

Customer-related question • Are customer characteristics linked to the occurrence of reclamations?

Conformance-related question • If the prescribed process flow is not followed, will this influence whether or not a reclamation occurs?

Conclusion • Proposed a framework to correlate business process characteristics • Useful for operational support (prediction and recommendation) and process improvement • Many well-studied problems are just ad-hoc instances for our framework, such as: • Decision Point Analysis • Prediction of the Remaining Time to the end of instance • Decision Support Systems to reduce costs, risks, nonconformance, global execution time • Implemented in Pro. M, Customizable and Extendible

? Join the demo session on Wednesday from 16. 00 to 17. 30!!