Ontologies And Process Mining
-
Upload
george-varvaressos -
Category
Business
-
view
1.165 -
download
4
description
Transcript of Ontologies And Process Mining
George [email protected]
Business Process Mining © 2009
Ontologies and their use in Semantic Process Mining
Semantic Process Mining
•Based on Ontologies
•How do we discover Ontologies?
•Start with OOBE Business Patterns®
Business Process Mining © 2009
OOBE Business Patterns
•Process Business Objects
•Entity Business Objects
•Event Business Objects
Business Process Mining © 2009
Process Business Objects
•Business Verbs
•Represent Processes
Business Process Mining © 2009
Process Business Objects• Seven Process Business Objects from which all others can be defined
Business Process Mining © 2009
Process BusinessObject
Developing Procuring Producing Promoting Selling Delivering Billing
Entity Business Objects
•Business Nouns
•People, Places & Things
•Resources, Players & Constructs
Business Process Mining © 2009
Entity Business Objects
Business Process Mining © 2009
EntityBusinessObject
Resource Product or Service Party Transfer Agreement Plan Location
Entity Object - Agreement
Business Process Mining © 2009
Agreement
Compensation Agreement
EmploymentAgreement
Risk Assumption
Contract
Insurance Policy Lease Guarantee
QuoteSupplier Contract Licence Policy
Legislation
Entity Object – Business Party
Business Process Mining © 2009
Business
Party
Individual
Consumer
Beneficiary Buyer Payee Recipient
Employee
Direct
Permanent
Full Time
Part Time
Temporary
Seconded
Organisation
Supplier Owner Employer
Regulator
Event Business Objects
•Business Events
•Business Time, Product & Life Cycle events
•Event Business Objects are triggers for or results of steps in Business Process Objects
Business Process Mining © 2009
Event Business Objects• Seven Event Business Objects from which all others can be defined
Business Process Mining © 2009
EventBusinessObject
Calendar Time Clock Time Business
TimeElapsed
TimeEnvironment
Time Bound Occurrence
Business Objects and Ontologies • Semantic Business Process Mining is based on Ontologies
http://www.ip-super.org
• OOBE Business Patterns® create a framework for the Ontologieshttp://www.promendo.com
• Incorporate Business Patterns into WSML (ontologies)http://www.wsmostudio.org
• Ontologies then annotate process mining data (ProM)http://prom.win.tue.nl/tools/prom/
• Patterns are important to ensure that nothing is missed
Business Process Mining © 2009
Process_ID EventName EventType Timestamp Originator Type Server Outcome
33116 VRU start 1/01/1999 0:00:31 No_Server PS No_Server Hang
33116 VRU complete 1/01/1999 0:00:36 No_Server PS No_Server Hang
33117 VRU start 1/01/1999 0:34:12 No_Server PS No_Server Hang
33117 VRU complete 1/01/1999 0:34:23 No_Server PS No_Server Hang
33118 VRU start 1/01/1999 6:55:20 Michael PS Michael Agent
33118 VRU complete 1/01/1999 6:55:26 Michael PS Michael Agent
33118 Regular_Service start 1/01/1999 6:55:44 Michael PS Michael Agent
33118 Regular_Service complete 1/01/1999 6:56:37 Michael PS Michael Agent
33119 VRU start 1/01/1999 7:41:16 Jack PS Jack Agent
33119 VRU complete 1/01/1999 7:41:26 Jack PS Jack Agent
33119 Regular_Service start 1/01/1999 7:41:27 Jack PS Jack Agent
33119 Regular_Service complete 1/01/1999 7:44:53 Jack PS Jack Agent
Process Discovery
Example of an ontology candidate
Business Process Mining © 2009
Semantic Data
Business Process Mining © 2009
Operations Dept manages human resources- by team
StatisticalData
Business Process Mining © 2009
Analyse performance by task & abstraction
Semantic Data
Business Process Mining © 2009
Payroll Deptmanages human resources by – Business Party
Statistical Data
Business Process Mining © 2009
Analyse performance using ‘Business Party’ abstraction
Summary•Ontologies are needed for Semantic Process Mining
•Business Patterns provide a template when you need to create an Ontology
•Ontologies provide different perspectives for the same data
•Same total result (Time & Frequency) in all Ontologies – just a different view
Business Process Mining © 2009