Knowledge management and business process management
Post on 11-May-2015
4533 Views
Preview:
DESCRIPTION
Transcript
© 2005
A framework for the improvement of knowledge intense business processesPeter Dalmaris
13-12-2005Room BC412Department of Industrial and Systems EngineeringPolytechnic University of Hong Kong
© 2005
What is the KBPI framework?
Knowledge-Based Process Improvement
On
tolo
gy
Me
tho
do
log
y
Conclusions
Questions
Introduction
Ep
iste
mo
log
y
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.
© 2005
What is the KBPI framework?Knowledge-Based Process Improvement
On
tolo
gy
Me
tho
do
log
y
Conclusions
Questions
Introduction
Ep
iste
mo
log
y
© 2005
How does it work? Targets knowledge-intense business
processes I.e. Loan approvement or R&D
processes Analyses its current knowledge-related
attributes Identifies areas of possible improvement Proposes a plan for improving
performance by improving the management of process knowledge
On
tolo
gy
Me
tho
do
log
y
Conclusions
Questions
Introduction
Ep
iste
mo
log
y
© 2005
What is a knowledge-intense business process?
On
tolo
gy
Me
tho
do
log
y
Conclusions
Questions
Introduction
Ep
iste
mo
log
y
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.
© 2005
Agenda: Discuss the components of the KBPI
1.EPISTEMOLOGY
2.ONTOLOGY
3.METHODOLOGY
© 2005
Why involve epistemology?
What is knowledge? Where is knowledge? How is knowledge created? What about data and information?
LEVEL 1: EPISTEMOLOGY
On
tolo
gy
Me
tho
do
log
y
Conclusions
Questions
Introduction
Ep
iste
mo
log
y
One needs to understand knowledge before speaking about knowledge.
Epistemology is the branch of philosophy that studies knowledge. It answers the basic questions:
© 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
On
tolo
gy
Me
tho
do
log
y
Conclusions
Questions
Introduction
Ep
iste
mo
log
y
© 2005
Understanding knowledge. What is it?
My definition:
Knowledge is solutions to problems
Heavily influenced by Karl Popper’s evolutionary epistemology
LEVEL 1: EPISTEMOLOGY
On
tolo
gy
Me
tho
do
log
y
Conclusions
Questions
Introduction
Ep
iste
mo
log
y
Problems drive knowledge creation knowledge consists of the solutions that drove its creation!
© 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/hVisibility:9.00 kmDewpoint:14°Barometer: UnknownSunrise:6:36Sunset17:40
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
On
tolo
gy
Me
tho
do
log
y
Conclusions
Questions
Introduction
Ep
iste
mo
log
y
…in all cases, knowledge, information and data must be considered in context.
© 2005
Understanding knowledge. Where is it?
Popper proposed 3 ontological worlds of human experience:
LEVEL 1: EPISTEMOLOGY
On
tolo
gy
Me
tho
do
log
y
Conclusions
Questions
Introduction
Ep
iste
mo
log
y
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
© 2005
Understanding knowledge. Where is it?
Diagram used with permission from Dr Joe Firestone, © 2004 KMCI
Popper’s 3 ontological worlds
LEVEL 1: EPISTEMOLOGY
On
tolo
gy
Me
tho
do
log
y
Conclusions
Questions
Introduction
Ep
iste
mo
log
y
© 2005
Understanding knowledge. How is it created?
Popper’s tetradic schema
P: a problem propositionTT: a tentative theory (solution)EE: error elimination (finding problems with the P and the TT)
LEVEL 1: EPISTEMOLOGY
On
tolo
gy
Me
tho
do
log
y
Conclusions
Questions
Introduction
Ep
iste
mo
log
y
© 2005
Understanding knowledge. No knowledge is perfectPopper’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
On
tolo
gy
Me
tho
do
log
y
Conclusions
Questions
Introduction
Ep
iste
mo
log
y
© 2005
Understanding knowledge. Data and information.
Data
Information
KnowledgeTOO ABSTRACT – TOO GENERAL
LEVEL 1: EPISTEMOLOGY
On
tolo
gy
Me
tho
do
log
y
Conclusions
Questions
Introduction
Ep
iste
mo
log
y
Problem contextSensors
SIGNALS
World Knowledge KnowledgeInformation
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
Men
tal a
ctio
nP
hysi
cal a
ctio
n
aacaataaca gcagtgagaa tgaacgcact taaataaaag ctcgtgtcta
Data
© 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
OrganisationalLearning
OrganisationalMemory
OrganisationalKnowledge
commits
createssu
pports
PROBLEM
trig
ger
s
On
tolo
gy
Me
tho
do
log
y
Conclusions
Questions
Introduction
Ep
iste
mo
log
yE
pis
tem
olo
gy
© 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
On
tolo
gy
Me
tho
do
log
y
Conclusions
Questions
Introduction
Ep
iste
mo
log
yE
pis
tem
olo
gy
© 2005
Discuss the components of the KBPI
1.EPISTEMOLOGY
2.ONTOLOGY
3.METHODOLOGY
© 2005
The business process ontology. What is ontology?
LEVEL 2: ONTOLOGY
Introduction
Ep
iste
mo
log
yE
pis
tem
olo
gy
Me
tho
do
log
y
Conclusions
Questions
On
tolo
gy
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)
© 2005
Difference between taxonomy and ontology.
LEVEL 2: ONTOLOGY
Introduction
Ep
iste
mo
log
yE
pis
tem
olo
gy
Me
tho
do
log
y
Conclusions
Questions
On
tolo
gy
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
© 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
Ep
iste
mo
log
yE
pis
tem
olo
gy
Me
tho
do
log
y
Conclusions
Questions
On
tolo
gy
I needed a way to formally describe knowledge-intense business processes.
© 2005
Business process ontology. Current revision.
LEVEL 2: ONTOLOGY
Introduction
Me
tho
do
log
y
Conclusions
Questions
Ep
iste
mo
log
yE
pis
tem
olo
gy
On
tolo
gy
© 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
Me
tho
do
log
y
Conclusions
Questions
Ep
iste
mo
log
yE
pis
tem
olo
gy
On
tolo
gy
© 2005
Abstract classes
Knowledge Object types Structural Functional Environmental Etc.
Knowledge Process types Transaction Types Containers and Media Medium Types
Used in support of the normal classes.
LEVEL 2: ONTOLOGY
Introduction
Ep
iste
mo
log
yM
eth
od
olo
gy
Conclusions
Questions
Ep
iste
mo
log
yE
pis
tem
olo
gy
On
tolo
gy
© 2005
Discuss the components of the KBPI
1.EPISTEMOLOGY
2.ONTOLOGY
3.METHODOLOGY
© 2005
Why methodology?
LEVEL 3: METHODOLOGY
Introduction
Ep
iste
mo
log
yO
nto
log
y
Conclusions
Questions
Ep
iste
mo
log
yE
pis
tem
olo
gy
Ep
iste
mo
log
yO
nto
log
yM
eth
od
olo
gy
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.
© 2005
The KBPI method
LEVEL 3: METHODOLOGY
Introduction
Ep
iste
mo
log
yO
nto
log
y
Conclusions
Questions
Ep
iste
mo
log
yE
pis
tem
olo
gy
Ep
iste
mo
log
yO
nto
log
yM
eth
od
olo
gy
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
© 2005
Audit procedure
LEVEL 3: METHODOLOGY
Introduction
Ep
iste
mo
log
yO
nto
log
y
Conclusions
Questions
Ep
iste
mo
log
yO
nto
log
yE
pis
tem
olo
gy
On
tolo
gy
Ep
iste
mo
log
yE
pis
tem
olo
gy
Ep
iste
mo
log
yO
nto
log
yM
eth
od
olo
gy
Audit
© 2005
Analysis: two levels
LEVEL 3: METHODOLOGY
Introduction
Ep
iste
mo
log
yO
nto
log
y
Conclusions
Questions
Ep
iste
mo
log
yO
nto
log
yE
pis
tem
olo
gy
Ep
iste
mo
log
yE
pis
tem
olo
gy
On
tolo
gy
Me
tho
do
log
y
© 2005
Function level procedureF1
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.
Op
era
tion
s o
n F
un
ctio
n c
lass
inst
ance
sO
pe
ratio
ns
on
oth
er
cla
ss in
sta
nce
sE
rro
r d
isco
very
an
d so
lutio
n d
esi
gn
LEVEL 3: METHODOLOGY
Ep
iste
mo
log
yO
nto
log
y
Conclusions
Questions
Ep
iste
mo
log
yO
nto
log
yE
pis
tem
olo
gy
Ep
iste
mo
log
yE
pis
tem
olo
gy
On
tolo
gy
Me
tho
do
log
y
© 2005
Process level procedureLEVEL 3: METHODOLOGY
Ep
iste
mo
log
yO
nto
log
y
Conclusions
Questions
Ep
iste
mo
log
yO
nto
log
yE
pis
tem
olo
gy
Ep
iste
mo
log
yE
pis
tem
olo
gy
On
tolo
gy
Me
tho
do
log
y
KP1
KP2
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:
Ope
ratio
ns o
n K
now
ledg
e P
ath
clas
s in
stan
ces
Ope
ratio
ns o
n ot
her
clas
s in
stan
ces
Err
or d
isco
very
and
so
lutio
n de
sign
KP3
KP4
© 2005
Tools usedLEVEL 3: METHODOLOGY
Ep
iste
mo
log
yO
nto
log
y
Conclusions
Questions
Ep
iste
mo
log
yO
nto
log
yE
pis
tem
olo
gy
Ep
iste
mo
log
yE
pis
tem
olo
gy
On
tolo
gy
Me
tho
do
log
y
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
© 2005
Tools: Protégé ontology editorLEVEL 3: METHODOLOGY
Introduction
Ep
iste
mo
log
yO
nto
log
y
Conclusions
Questions
Ep
iste
mo
log
yO
nto
log
yE
pis
tem
olo
gy
On
tolo
gy
Ep
iste
mo
log
yO
nto
log
yE
pis
tem
olo
gy
Ep
iste
mo
log
yE
pis
tem
olo
gy
On
tolo
gy
Me
tho
do
log
y
© 2005
Tools: Graphic modeller
LEVEL 3: METHODOLOGY
Introduction
Ep
iste
mo
log
yO
nto
log
y
Conclusions
Questions
Ep
iste
mo
log
yO
nto
log
yE
pis
tem
olo
gy
On
tolo
gy
Ep
iste
mo
log
yO
nto
log
yE
pis
tem
olo
gy
On
tolo
gy
Ep
iste
mo
log
yE
pis
tem
olo
gy
Ep
iste
mo
log
yO
nto
log
yM
eth
od
olo
gy
© 2005
Concluding remark 1: Innovation Innovation is in the application of an
evolutionary epistemology. I now have a better idea of what knowledge
is.
Innovation is in the use of an ontology in business processes. I can now describe knowledge-intense
business processes using a formal language
Introduction
Ep
iste
mo
log
yO
nto
log
yM
eth
od
olo
gy
Conclusions
Questions
Ep
iste
mo
log
yO
nto
log
yE
pis
tem
olo
gy
On
tolo
gy
Ep
iste
mo
log
yO
nto
log
yE
pis
tem
olo
gy
On
tolo
gy
Ep
iste
mo
log
yO
nto
log
yE
pis
tem
olo
gy
Ep
iste
mo
log
yE
pis
tem
olo
gy
On
tolo
gy
Me
tho
do
log
y
© 2005
Concluding remark 2: Business ontology The business ontology will be
improved with time. An improved business ontology will
allow for a more precise definition of the business processThis will allow for more accurate
analysis and tentative solutions (improvement recommendations)
Introduction
Ep
iste
mo
log
yO
nto
log
yM
eth
od
olo
gy
Questions
Ep
iste
mo
log
yO
nto
log
yM
eth
od
olo
gy
Ep
iste
mo
log
yO
nto
log
yE
pis
tem
olo
gy
On
tolo
gy
Ep
iste
mo
log
yO
nto
log
yE
pis
tem
olo
gy
On
tolo
gy
Ep
iste
mo
log
yO
nto
log
yE
pis
tem
olo
gy
Ep
iste
mo
log
yE
pis
tem
olo
gy
On
tolo
gy
Conclusions
Me
tho
do
log
y
© 2005
Concluding remark 3: Tools
Much of the methodology can be automated/facilitated with appropriate use of tools.
The KBPI based on the use of Protégé is a first step towards a knowledge engineering software suite.
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).
The third step is a secret.
Introduction
Ep
iste
mo
log
yO
nto
log
yM
eth
od
olo
gy
Questions
Ep
iste
mo
log
yO
nto
log
yM
eth
od
olo
gy
Ep
iste
mo
log
yO
nto
log
yM
eth
od
olo
gy
Ep
iste
mo
log
yO
nto
log
yE
pis
tem
olo
gy
On
tolo
gy
Ep
iste
mo
log
yO
nto
log
yE
pis
tem
olo
gy
On
tolo
gy
Ep
iste
mo
log
yO
nto
log
yE
pis
tem
olo
gy
Ep
iste
mo
log
yE
pis
tem
olo
gy
On
tolo
gy
Conclusions
Me
tho
do
log
y
© 2005
Thanks to:
Socrates of Athens, and Sir Karl R. Popper for their clarity and wisdom
Dr Eric Tsui for his advice over the years and invitation to PolyU
Dr Ken Dovey (University of Technology, Syndey), Dr Bill Hall (Tenix Defence, Melbourne), Dr Bob Smith (Tall Tree Labs)
My dissertation examiners for their valuable critique towards eliminating my errors.
Introduction
Ep
iste
mo
log
yO
nto
log
yM
eth
od
olo
gy
Questions
Ep
iste
mo
log
yO
nto
log
yM
eth
od
olo
gy
Ep
iste
mo
log
yO
nto
log
yM
eth
od
olo
gy
Ep
iste
mo
log
yO
nto
log
yM
eth
od
olo
gy
Ep
iste
mo
log
yO
nto
log
yE
pis
tem
olo
gy
On
tolo
gy
Ep
iste
mo
log
yO
nto
log
yE
pis
tem
olo
gy
On
tolo
gy
Ep
iste
mo
log
yO
nto
log
yE
pis
tem
olo
gy
Ep
iste
mo
log
yE
pis
tem
olo
gy
On
tolo
gy
Conclusions
Me
tho
do
log
y
© 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
Ep
iste
mo
log
yO
nto
log
yM
eth
od
olo
gy
Ep
iste
mo
log
yO
nto
log
yM
eth
od
olo
gy
Ep
iste
mo
log
yO
nto
log
yM
eth
od
olo
gy
Ep
iste
mo
log
yO
nto
log
yM
eth
od
olo
gy
Ep
iste
mo
log
yO
nto
log
yM
eth
od
olo
gy
Ep
iste
mo
log
yO
nto
log
yE
pis
tem
olo
gy
On
tolo
gy
Ep
iste
mo
log
yO
nto
log
yE
pis
tem
olo
gy
On
tolo
gy
Ep
iste
mo
log
yO
nto
log
yE
pis
tem
olo
gy
Ep
iste
mo
log
yE
pis
tem
olo
gy
On
tolo
gy
Conclusions
Questions
Me
tho
do
log
y
top related