Conformance Checking of Processes Based on Monitoring Real...
Transcript of Conformance Checking of Processes Based on Monitoring Real...
Conformance Checking of Processes Based on Monitoring Real Behavior
Ronny Mans
Department of Technology Management PAGE 124-4-2009
Conformance Checking
Prof. Wil van der AalstAnne
Department of Technology Management PAGE 224-4-2009
Motivation
How to check whether interacting parties stick to what they agreed upon?
insurance company
car rental agency
hotel chain
travel agency
airline
Department of Technology Management PAGE 324-4-2009
Outline
• Introduction Process Mining• Conformance Checking
• Fitness• Appropriateness
• Application to Web Services• Feasibility Study• Conclusion
Department of Technology Management PAGE 424-4-2009
1. Introduction Process Mining
• analyze run-time data• possibly relate to existing
models and/or requirements• process mining!
processdesign
implementation/configuration
processenactment
diagnosisprocesscontrol
Department of Technology Management PAGE 524-4-2009
2. Conformance Checking
Objectives:•quantitatively measure conformance (i.e., metrics)•locate deviations
Conformance?
Department of Technology Management PAGE 624-4-2009
2.1 Conformance Checking - Fitness
Department of Technology Management PAGE 724-4-2009
2.1 Conformance Checking - Fitness
Department of Technology Management PAGE 824-4-2009
2.1 Measuring fitness: Log replay analysis
missing tokensremaining tokens
= 0= 0
= 0= 1
consumed tokensproduced tokens
Department of Technology Management PAGE 924-4-2009
2.1 Measuring fitness: Log replay analysis
= 1= 2
= 0= 0
missing tokensremaining tokens
consumed tokensproduced tokens
Department of Technology Management PAGE 1024-4-2009
2.1 Measuring fitness: Log replay analysis
= 2= 4
= 0= 0
missing tokensremaining tokens
consumed tokensproduced tokens
Department of Technology Management PAGE 1124-4-2009
2.1 Measuring fitness: Log replay analysis
= 3= 5
= 0= 0
missing tokensremaining tokens
consumed tokensproduced tokens
Department of Technology Management PAGE 1224-4-2009
2.1 Measuring fitness: Log replay analysis
= 5= 6
= 0= 0
missing tokensremaining tokens
consumed tokensproduced tokens
Department of Technology Management PAGE 1324-4-2009
2.1 Measuring fitness: Log replay analysis
= 6= 7
= 0= 0
missing tokensremaining tokens
consumed tokensproduced tokens
Department of Technology Management PAGE 1424-4-2009
2.1 Measuring fitness: Log replay analysis
= 7= 7
= 0= 0
missing tokensremaining tokens
consumed tokensproduced tokens
Department of Technology Management PAGE 1524-4-2009
2.1 Measuring fitness: Log replay analysis
= 9= 9
= 0= 0
missing tokensremaining tokens
consumed tokensproduced tokens
Department of Technology Management PAGE 1624-4-2009
2.1 Measuring fitness: Log replay analysis
= 9= 9
= 0= 0
missing tokensremaining tokens
consumed tokensproduced tokens
Department of Technology Management PAGE 1724-4-2009
2.1 Measuring fitness: Log replay analysis
= 0= 1
= 0= 0
missing tokensremaining tokens
consumed tokensproduced tokens
Department of Technology Management PAGE 1824-4-2009
2.1 Measuring fitness: Log replay analysis
= 1= 2
= 0= 0
missing tokensremaining tokens
consumed tokensproduced tokens
Department of Technology Management PAGE 1924-4-2009
2.1 Measuring fitness: Log replay analysis
= 2= 4
= 1= 0
missing tokensremaining tokens
consumed tokensproduced tokens
Department of Technology Management PAGE 2024-4-2009
2.1 Measuring fitness: Log replay analysis
= 3= 5
= 1= 0
missing tokensremaining tokens
consumed tokensproduced tokens
Department of Technology Management PAGE 2124-4-2009
2.1 Measuring fitness: Log replay analysis
= 4= 6
= 1= 0
missing tokensremaining tokens
consumed tokensproduced tokens
Department of Technology Management PAGE 2224-4-2009
2.1 Measuring fitness: Log replay analysis
= 6= 7
= 1= 0
missing tokensremaining tokens
consumed tokensproduced tokens
Department of Technology Management PAGE 2324-4-2009
2.1 Measuring fitness: Log replay analysis
= 7= 8
= 1= 0
missing tokensremaining tokens
consumed tokensproduced tokens
Department of Technology Management PAGE 2424-4-2009
2.1 Measuring fitness: Log replay analysis
= 8= 8
= 1= 1
missing tokensremaining tokens
consumed tokensproduced tokens
Department of Technology Management PAGE 2524-4-2009
2.1 Measuring fitness: Log replay analysis
= 8= 8
= 1= 1
missing tokensremaining tokens
consumed tokensproduced tokens
Department of Technology Management PAGE 2624-4-2009
2.1 Measuring fitness: Log replay analysis
Department of Technology Management PAGE 2724-4-2009
2.1 Measuring fitness: Log replay analysis
f = 1.0 f ≈ 0.540f ≈ 0.955
Department of Technology Management PAGE 2824-4-2009
2.2 Conformance Checking - Appropriateness
100 % fitnessbut not sufficiently specific from behavioral point of view.
100 % fitnessbut not represented instructurally suitable way.
Department of Technology Management PAGE 2924-4-2009
3. Application to Web Services
• Service-oriented systems are composed of services that are:• independently developed and operated• interact with one another exclusively through
message exchanges➡ Expected behavior may deviate
• e.g., service receives reply of the wrong type, messages may be received in the wrong order, ...
Question: “Do all services in a service-oriented system operate as expected?”
Department of Technology Management PAGE 3024-4-2009
3. Application to Web Services - Approach
ServicesServicesabstractBPEL
process
Petri netmodel
1.
SOAPMessages
Event Log
Services
Monitoring /Correlation
Conformance?
2.
ConformanceChecking!
Translation
Department of Technology Management PAGE 3124-4-2009
3.1 From BPEL to Petri Nets
Department of Technology Management PAGE 3224-4-2009
3.1 From BPELto Petri Nets
Translation isfeature-complete:
➡ maps all BPELaspects on a Petri-net model
C. Ouyang, W.M.P. van der Aalst, S. Breutel, M. Dumas, A.H.M. ter Hofstede, and H.M.W. Verbeek. Formal Semantics and Analysis of Control Flow in WS-BPEL (Revised version). BPM Center Report BPM-05-15, BPMcenter.org, 2005.
Department of Technology Management PAGE 3324-4-2009
3.2 Monitoring and Correlating Messages
From SOAP messages to Event log:• Process instance (PI)• Label denoting the BPEL communication
action (MT)
Department of Technology Management PAGE 3424-4-2009
Service Behavior
4. Feasibility Study - Overview
abstractBPEL
process
BPEL2PNML+
WofBPEL
WF-net
1.
SOAPMessages
ConformanceChecker(ProM)
3.
MXMLEvent Log
OracleBPELServer
OracleBPELServer
Conformance?
2.
TCP tunneling +Conversion
Department of Technology Management PAGE 3524-4-2009
4.1 From BPEL to WF-nets (1)
• Translation: BPEL2PNML
Department of Technology Management PAGE 3624-4-2009
4.1 From BPEL to WF-nets (2)
• Reduction: WofBPEL
Department of Technology Management PAGE 3724-4-2009
4.2 Observing Service Behavior
Implementation via Oracle BPEL:• Supplier service• Customer service
➡could have been implementedon any other platform! (e.g., Java, .NET, etc.)
Service Behavior
SOAPMessages
MXMLEvent Log
OracleBPELServer
OracleBPELServer
TCP tunneling +Conversion
Result: Observation of one out of many possible communication scenarios
Department of Technology Management PAGE 3824-4-2009
4.3 Conformance Checking
• Desirable and undesirable scenarios for the supplier service execution
Department of Technology Management PAGE 3924-4-2009
4.3 Conformance Checking
• Analysis scenario 8: (orderResponse)
Department of Technology Management PAGE 4024-4-2009
4. Conclusion
• Conformance checking particularly relevant in a service-based environment due to the independence of services
• Feasible to check conformance of an abstract BPEL process using existing tool chain (BPEL2PNML, WofBPEL, ProM Conformance Checker)
W.M.P. van der Aalst, M. Dumas, C. Ouyang, A. Rozinat, and H.M.W. Verbeek. Choreography Conformance Checking: An Approach based on BPEL andPetri Nets (extended version). BPM Center Report BPM-05-25, BPMcenter.org, 2005.
Department of Technology Management PAGE 4124-4-2009
Thank you very much!
Questions?