A knowledge engineering framework for improving business processe

The presentation is based on a paper by the presenter (published in the Business Process Management Journal). It reports on the design of a field tested framework for improving knowledge-intensive business processes. It addresses a gap in the area of knowledge-based process improvement identified via a literature review. It also addresses weaknesses and shortcomings related to the knowledge aspects of business processes of established process improvement methodologies such as TQM and Six Sigma. The proposed framework is based on Karl Popper’s evolutionary epistemology of Objective Knowledge and integrates concepts and techniques typically applied in business process management and knowledge management. Through the conduct of three case studies, the improvement framework is applied and evaluated in real organisational settings. The lessons learned from these case studies and practical applications of the framework are discussed.

Transcript

1. © 2005 A framework for the improvement of knowledge intense business processes Peter Dalmaris 13-12-2005 Room BC412 Department of Industrial and Systems Engineering Polytechnic University of Hong Kong
2. © 2005 What is the KBPI framework? Knowledge-Based Process Improvement Ontology Methodology Conclusions Questions Introduction Epistemology The KBPI framework is a tool for the improvement of knowledge-intense business processes. • It is based on Karl Popper’s evolutionary epistemology; this provides the theoretical foundations. • It uses a business process ontology; this provides a language for describing business processes. • It applies an improvement methodology; this provides the practical steps of improvement.
3. © 2005 What is the KBPI framework? Knowledge-Based Process Improvement Ontology Methodology Conclusions Questions Introduction Epistemology
4. © 2005 How does it work? n  Targets knowledge-intense business processes ¨ I.e. Loan approvement or R&D processes n  Analyses its current knowledge-related attributes n  Identifies areas of possible improvement n  Proposes a plan for improving performance by improving the management of process knowledge Ontology Methodology Conclusions Questions Introduction Epistemology
5. © 2005 What is a knowledge-intense business process? Ontology Methodology Conclusions Questions Introduction Epistemology Eppler, DMJ, Seifried, PM & Ropnack, A 1999, 'Improving Knowledge Intensive Processes through an Enterprise Knowledge Medium', SIGCPR'99, ACM, New Orleans, USA, pp. 222-30 Process complexity: High in process steps, involved agents, interdependency, process dynamic. Process intensity: Strong in contingency, decision scope, agent innovation, half-life, agent impact, learning time.
6. © 2005 Agenda: Discuss the components of the KBPI 1. EPISTEMOLOGY 2. ONTOLOGY 3. METHODOLOGY
7. © 2005 Why involve epistemology? n  What is knowledge? n  Where is knowledge? n  How is knowledge created? n  What about data and information? LEVEL 1: EPISTEMOLOGY Ontology Methodology Conclusions Questions Introduction Epistemology One needs to understand knowledge before speaking about knowledge. Epistemology is the branch of philosophy that studies knowledge. It answers the basic questions:
8. © 2005 Understanding knowledge: what is it? Justified true belief. (Goldman, Nonaka and Takeuchi) Understanding based on experience. (James 1907) Knowledge can be thought of as the body of understandings, generalizations, and abstractions that we carry with us on a permanent or semi-permanent basis and apply to interpret and manage the world around us ... we will consider knowledge to be the collection of mental units of all kinds that provides us with understanding and insights. (Wiig 1998) TOO ABSTRACT – TOO GENERAL LEVEL 1: EPISTEMOLOGY Ontology Methodology Conclusions Questions Introduction Epistemology
9. © 2005 Understanding knowledge. What is it? My definition: Knowledge is solutions to problems Heavily influenced by Karl Popper’s evolutionary epistemology LEVEL 1: EPISTEMOLOGY Ontology Methodology Conclusions Questions Introduction Epistemology Problems drive knowledge creation à knowledge consists of the solutions that drove its creation!
10. © 2005 Understanding knowledge. What is it? Knowledge is solutions to problems. Therefore: 1. Enable POP in your Gmail account. 2. Open Netscape Mail 7.x. 3. Click 'Edit,' and select 'Mail & Newsgroups Account Settings...' 4. Click 'Add Account...,' and click 'OK.' Knowledge Humidity:57% Wind:NNW/14 km/h Visibility:9.00 km Dewpoint:14° Barometer: Unknown Sunrise:6:36 Sunset17:40 Information 1 gctgtcagaa aacaataaca gcagtgagaa tgaacgcact taaataaaag ctcgtgtcta 61 gagtctctcc ttttataggc ctttcatgca aataaagaat tcaaaatatc cagctctgat 121 tgggcaatgt gttagtgacg catacatgta aaatagcctt caccttattt cctttctaat 181 tggttggctc gtcaaagaac aattttaacc aatcaaattg cgcctttcac aattctaccg 241 atgactataa ctagcttctt attcctccat cgagcccatt ctttttcttt attcagtgga 301 ttgttagttc ttctgctgtt aggaagccac tatgtctgga cgtggaaagc aaggcggcaa Data LEVEL 1: EPISTEMOLOGY Ontology Methodology Conclusions Questions Introduction Epistemology …in all cases, knowledge, information and data must be considered in context.
11. © 2005 Understanding knowledge. Where is it? Popper proposed 3 ontological worlds of human experience: LEVEL 1: EPISTEMOLOGY Ontology Methodology Conclusions Questions Introduction Epistemology WORLD 1: The world of material objects. Trees, chairs, our bodies belong here. WORLD 2: The world of mental states. Beliefs, dispositions, pleasure and dislikes belong here. WORLD 3: The world of books, words, statements and other such immaterial human creations. Theories, arguments, symphonies and paintings belong here. Immaterial but objective Immaterial but subjective Material
12. © 2005 Understanding knowledge. Where is it? Diagram used with permission from Dr Joe Firestone, © 2004 KMCI Popper’s 3 ontological worlds LEVEL 1: EPISTEMOLOGY Ontology Methodology Conclusions Questions Introduction Epistemology
13. © 2005 Understanding knowledge. How is it created? Popper’s tetradic schema P: a problem proposition TT: a tentative theory (solution) EE: error elimination (finding problems with the P and the TT) LEVEL 1: EPISTEMOLOGY Ontology Methodology Conclusions Questions Introduction Epistemology
14. © 2005 Understanding knowledge. No knowledge is perfect Popper’s tetradic schema is based on the tradition of fallibilism. Fallibilism: The idea that while universal knowledge claims cannot be confirmed or verified by empirical testing, they can be falsified, but also not with certainty. Firestone, J., McElroy, M., 2003, Key Issues in the New Knowledge Management, page 228 Socrates: All I know is that I know nothing Popper: There are no authoritative sources of knowledge, and no ‘source’ is particularly reliable. LEVEL 1: EPISTEMOLOGY Ontology Methodology Conclusions Questions Introduction Epistemology
15. © 2005 Understanding knowledge. Data and information. Data Information Knowledge TOO ABSTRACT – TOO GENERAL LEVEL 1: EPISTEMOLOGY Ontology Methodology Conclusions Questions Introduction Epistemology Problem context Sensors SIGNALS World Knowledge Knowledge Information World broadcasts purposeless and contextless signals Signals are captured by sensor(s) and turned into data Information is constructed from the data Information is processed using knowledge to enable action or knowledge creation Action is taken on the world New knowledge is applied Mental action Physical action aacaataaca gcagtgagaa tgaacgcact taaataaaag ctcgtgtcta Data
16. © 2005 What is knowledge management? Knowledge management is the managerial activity charged with the responsibility of managing the organisational knowledge life- cycle in support of the organisation’s objectives and business processes. LEVEL 1: EPISTEMOLOGY Organisational Learning Organisational Memory Organisational Knowledge commits creates supports PROBLEM triggers Ontology Methodology Conclusions Questions Introduction Epistemology Epistemology
17. © 2005 Epistemological assumptions LEVEL 1: EPISTEMOLOGY Most knowledge useful to business processes can be objectified Knowledge can become separated from its creator Knowledge must be challenged relentlessly Fallibilism There is no perfect knowledge Less emphasis on the knower More emphasis on the knowledge objects (world 3) Of course, personal (world 2) knowledge are still very important Ontology Methodology Conclusions Questions Introduction Epistemology Epistemology
18. © 2005 Discuss the components of the KBPI 1. EPISTEMOLOGY 2. ONTOLOGY 3. METHODOLOGY
19. © 2005 The business process ontology. What is ontology? LEVEL 2: ONTOLOGY Introduction Epistemology Epistemology Methodology Conclusions Questions Ontology Tom Gruber, http://www-ksl.stanford.edu/kst/what-is-an-ontology.html • An ontology defines the vocabulary with which queries and assertions are exchanged among agents. • Ontological commitments are agreements to use the shared vocabulary in a coherent and consistent manner. • A commitment to a common ontology is a guarantee of consistency, but not completeness, with respect to queries and assertions using the vocabulary defined in the ontology. In Information Science, an ontology is the product of an attempt to formulate an exhaustive and rigorous conceptual schema about a domain. http://en.wikipedia.org/wiki/Ontology_(computer_science)
20. © 2005 Difference between taxonomy and ontology. LEVEL 2: ONTOLOGY Introduction Epistemology Epistemology Methodology Conclusions Questions Ontology Taxonomy is the science of classification – or a classification In Information Science, an ontology is the product of an attempt to formulate an exhaustive and rigorous conceptual schema about a domain. http://en.wikipedia.org/wiki/Ontology_(computer_science) http://en.wikipedia.org/wiki/Taxonomy
21. © 2005 The business process ontology. Why ontology? LEVEL 2: ONTOLOGY 1.  An ontology provides a formal conceptual schema/ model of a given domain. 2.  We need a formal description of a business process before we can do any work 3.  We need a vocabulary and syntax before we can communicate. Introduction Epistemology Epistemology Methodology Conclusions Questions Ontology I needed a way to formally describe knowledge-intense business processes.
22. © 2005 Business process ontology. Current revision. LEVEL 2: ONTOLOGY Introduction Methodology Conclusions Questions Epistemology Epistemology Ontology
23. © 2005 Definitions (partial) Knowledge object: A knowledge object is knowledge that has been objectified and exists in world three. In organisations, such knowledge objects are strategic plans, product specifications, marketing ideas etc. Knowledge Path: A Knowledge Path is concerned with the set of functions and their sequence of execution that perform some desired knowledge processing on a knowledge object. This knowledge processing may be an intermediate or a final deliverable of a knowledge-intensive business process. Knowledge Transaction: Knowledge transactions refer to the exchange of knowledge objects between actors within a business process. The word 'actor' is used here in its broad sense to mean humans or machines that can be receivers or transmitters of the knowledge objects. When a knowledge object is transferred from one actor to another, a transaction occurs. LEVEL 2: ONTOLOGY Introduction Methodology Conclusions Questions Epistemology Epistemology Ontology
24. © 2005 Abstract classes n  Knowledge Object types ¨  Structural ¨  Functional ¨  Environmental ¨  Etc. n  Knowledge Process types n  Transaction Types n  Containers and Media n  Medium Types Used in support of the normal classes. LEVEL 2: ONTOLOGY Introduction Epistemology Methodology Conclusions Questions Epistemology Epistemology Ontology
25. © 2005 Discuss the components of the KBPI 1. EPISTEMOLOGY 2. ONTOLOGY 3. METHODOLOGY
26. © 2005 Why methodology? LEVEL 3: METHODOLOGY Introduction Epistemology Ontology Conclusions Questions Epistemology Epistemology Epistemology Ontology Methodology I need a recipe of “how to improve a business process”. This recipe should tell me how to: 1.  Collect the data that describes the process 2.  Analyse the data 3.  Produce the results … all in a systematic and disciplined way.
27. © 2005 The KBPI method LEVEL 3: METHODOLOGY Introduction Epistemology Ontology Conclusions Questions Epistemology Epistemology Epistemology Ontology Methodology Knowledge Tools Knowledge Paths Knowledge Transactions Identify potential improvement areas ¦(desired process performance) Process Members Environment: constraints, policies, targets Audit: Probing, current state of the process (AS IS) Design: Result (AS COULD) Analysis: Improvement improvement configuration of process classes Functions Knowledge Objects Knowledge Transformations
28. © 2005 Audit procedure LEVEL 3: METHODOLOGY Introduction Epistemology Ontology Conclusions Questions Epistemology Ontology Epistemology Ontology Epistemology Epistemology Epistemology Ontology Methodology Audit
29. © 2005 Analysis: two levels LEVEL 3: METHODOLOGY Introduction Epistemology Ontology Conclusions Questions Epistemology Ontology Epistemology Epistemology Epistemology Ontology Methodology
30. © 2005 Function level procedure F1 F4 F3 F2 PM KO KT KX E KX F1: Find all knowledge intensive functions F2: Designate performance descriptors. F3: Determine current performance. F4: Determine desired performance. For each Function class instance: For each of Process member, Knowledge Object, Knowledge Transformation and Knowledge Tool class instances : KT: Define the Knowledge Tool instance. KO: Define the Knowledge Object instance. KX: Define the Knowledge Transformation instance. PM: Define the Process Member instance. Determine their Critical Knowledge Success Factors. For each of KT, KO, KX, PM, evaluate their current status and the impact of their performance on the Function performance. For each non-alignment: E: Find the likely causes. S: Design a possible solution. Operations on Function class instances Operations on other class instances Error discovery and solution design LEVEL 3: METHODOLOGY Epistemology Ontology Conclusions Questions Epistemology Ontology Epistemology Epistemology Epistemology Ontology Methodology
31. © 2005 Process level procedure LEVEL 3: METHODOLOGY Epistemology Ontology Conclusions Questions Epistemology Ontology Epistemology Epistemology Epistemology Ontology Methodology KP 1 KP 2 KT TR E KX KP1: Find all Knowledge Paths KP2: Designate performance descriptors. KP3: Determine current performance. KP4: Determine desired performance. For each Function class instance: For each of Knowledge Transaction and Knowledge Tool class instances : For each of KT, TR, evaluate their current status and the impact of their performance on the Knowledge Path performance. For each non-alignment: Operations on Knowledge Path class instances Operations on other class instances Error discovery and solution design KP 3 KP 4
32. © 2005 Tools used LEVEL 3: METHODOLOGY Epistemology Ontology Conclusions Questions Epistemology Ontology Epistemology Epistemology Epistemology Ontology Methodology 1.  Protégé, an open-source ontology editor •  For capturing and organising the process audit data. •  For enforcing the process ontology 2.  MS Visio •  For visualising the process model •  Uses standard BPML notation developed by BPMI.org •  For visualising some of the instances of the business ontology
33. © 2005 Tools: Protégé ontology editor LEVEL 3: METHODOLOGY Introduction Epistemology Ontology Conclusions Questions Epistemology Ontology Epistemology Ontology Epistemology Ontology Epistemology Epistemology Epistemology Ontology Methodology
34. © 2005 Tools: Graphic modeller LEVEL 3: METHODOLOGY Introduction Epistemology Ontology Conclusions Questions Epistemology Ontology Epistemology Ontology Epistemology Ontology Epistemology Ontology Epistemology Epistemology Epistemology Ontology Methodology
35. © 2005 Concluding remark 1: Innovation n  Innovation is in the application of an evolutionary epistemology. ¨ I now have a better idea of what knowledge is. n  Innovation is in the use of an ontology in business processes. ¨ I can now describe knowledge-intense business processes using a formal language Introduction Epistemology Ontology Methodology Conclusions Questions Epistemology Ontology Epistemology Ontology Epistemology Ontology Epistemology Ontology Epistemology Ontology Epistemology Epistemology Epistemology Ontology Methodology
36. © 2005 Concluding remark 2: Business ontology n  The business ontology will be improved with time. n  An improved business ontology will allow for a more precise definition of the business process ¨ This will allow for more accurate analysis and tentative solutions (improvement recommendations) Introduction Epistemology Ontology Methodology Questions Epistemology Ontology Methodology Epistemology Ontology Epistemology Ontology Epistemology Ontology Epistemology Ontology Epistemology Ontology Epistemology Epistemology Epistemology Ontology Conclusions Methodology
37. © 2005 Concluding remark 3: Tools n  Much of the methodology can be automated/ facilitated with appropriate use of tools. n  The KBPI based on the use of Protégé is a first step towards a knowledge engineering software suite. n  The second (small) step is the extension of Protégé to automate part of the analysis procedures. I am working on this now (in my spare time). n  The third step is a secret. Introduction Epistemology Ontology Methodology Questions Epistemology Ontology Methodology Epistemology Ontology Methodology Epistemology Ontology Epistemology Ontology Epistemology Ontology Epistemology Ontology Epistemology Ontology Epistemology Epistemology Epistemology Ontology Conclusions Methodology
38. © 2005 Thanks to: n  Socrates of Athens, and Sir Karl R. Popper for their clarity and wisdom n  Dr Eric Tsui for his advice over the years and invitation to PolyU n  Dr Ken Dovey (University of Technology, Syndey), Dr Bill Hall (Tenix Defence, Melbourne), Dr Bob Smith (Tall Tree Labs) n  My dissertation examiners for their valuable critique towards eliminating my errors. Introduction Epistemology Ontology Methodology Questions Epistemology Ontology Methodology Epistemology Ontology Methodology Epistemology Ontology Methodology Epistemology Ontology Epistemology Ontology Epistemology Ontology Epistemology Ontology Epistemology Ontology Epistemology Epistemology Epistemology Ontology Conclusions Methodology
39. © 2005 Questions Contact me: •  peter.dalmaris@futureshock.com.au •  94906537 (In Hong Kong until January 2) •  +61414685581 (In Sydney) •  Fax: +61 2 821 259 38 Introduction Epistemology Ontology Methodology Epistemology Ontology Methodology Epistemology Ontology Methodology Epistemology Ontology Methodology Epistemology Ontology Methodology Epistemology Ontology Epistemology Ontology Epistemology Ontology Epistemology Ontology Epistemology Ontology Epistemology Epistemology Epistemology Ontology Conclusions Questions Methodology