Building the knowledge map : an industrial case study

1
Building the knowledge map an industrial case
study
Suyeon Kim Euiho Suh Hyunseok Hwang
JOURNAL OF KNOWLEDGE MANAGEMENT, VOL. 7 2003
2
Contents

• Introduction
• Related works
• A conceptual framework
• Procedures of building the knowledge map
• Defining organizational knowledge
• Process map analysis
• Knowledge extraction
• Knowledge profiling
• Knowledge linking
• Knowledge map validation
• Case study P steel company
• Conclusion

3
1. Introduction(1/2)

• Knowledge management
• Many organizations currently engage in knowledge
  management in order to leverage knowledge both
  within their organization and externally to their
  shareholders and customers
• (Rubenstein-Montano, 2001)
• A significant and time consuming problem for
  knowledge-based system developers is how to
  efficiently elicit knowledge from experts and
  transform this elicited knowledge into a machine
  usable format (Walczak,1998)

Capturing and representing knowledge buried in
people and an organization are the fundamental
building blocks of knowledge management
implementation
4
1. Introduction(2/2)

• This paper provides
• Practical methodology
• Six-step procedure
• Knowledge map
• Several techniques roadmap
• A case study (lessons learned from an industrial
  case study)

5
2. Related works(1/5)

• Frame
• A powerful knowledge representation system that
  is accessible to both humans and machines.
• A collection of information and associated
  actions that represents a simple concept (Gordon,
  2000)
• Petri net
• Petri net can depict important knowledge about
  the structure and dynamic behavior of the system
• Rudas and Horvath (1997) used Petri net for
  process model entities and their evaluation
  knowledge

Petri net model
Transitions
Places
6
2. Related works(2/5)

• Semantic network
• Each node is specific knowledge and links show
  the interrelationship between knowledge.
• This means that they can also become a vehicle to
  archive company knowledge (Gordon, 2000).

7
2. Related works(3/5)

• Concept mapping
• Node-link structure (a concept map), in which
  nodes denote concepts, and links show the
  relationships between these concepts
• A type of structured conceptualization used by
  groups to develop a conceptual framework which
  can guide evaluation or planning (Trochim, 1989)

8
2. Related works(4/5)

• Concept mapping

9
2. Related works(5/5)

• Knowledge mapping
• The process, methods and tools for analyzing
  knowledge areas in order to discover features or
  meaning and to visualize these in a
  comprehensive, transparent form, such that the
  business- relevant features are clearly
  highlighted (Speel,1999)
• A navigation aid to explicit and tacit knowledge,
  illustrating how knowledge flows throughout an
  organization
• (Grey, 1999)
• Developing a knowledge map involves locating
  important knowledge in the organization and then
  publishing some sort of list or picture that
  shows where to find it
• (Davenport and Prusak, 1998)

10
3. A conceptual framework(1/3)

• Knowledge map is approach to represent explicit
  and tacit knowledge
• within an organization

ltA conceptual model of knowledge mapgt

• Diagram Graphical representation of knowledge
  (node, linkage)
• Node Rectangular object, Knowledge captured from
  business process
• Linkage Arrow between nodes implying
  relationships among knowledge
• Specification Descriptive representation of
  knowledge

11
3. A conceptual framework(2/3)

• Advantages from building knowledge map
• Formalization of all knowledge inventories in
  organization
• Perception of relationships between knowledge
• Efficient navigation of knowledge inventory
• Promotion of socialization / externalization of
  knowledge by connecting domain experts with
  knowledge explorers
• All captured knowledge can be summarized and
  abstracted through the knowledge map
• also gives a useful blueprint for implementing a
  knowledge management system (KMS)

12
3. A conceptual framework(3/3)
ltA knowledge management frameworkgt
The knowledge map plays a key role in the KM
project because it gives a knowledge profile
(knowledge warehouse), knowledge link (navigation
aids among knowledge), and expert finder
13
Question 1
J.H. Lee

• I agree that the knowledge map is very helpful
  for creating KMS.However, I wonder whether it is
  also helpful for maintaining KMS or not. Who
  maintains KMS (updates old knowledge and create
  new knowledge)? if every worker maintains their
  each knowledge, should they also understand the
  knowledge map fully?

Create, Update
KMS
Search, Query
Knowledge map Self construction
14
4. Procedures of buildingthe knowledge map(1/10)
ltProcedures of building the knowledge mapgt
15
4. Procedures of buildingthe knowledge map(2/10)

• Step.1 Defining organizational knowledge
• Defining knowledge ontology
• To provide a uniform, text-based intermediate
  representation of the knowledge types specific to
  a development effort that is comprehensible by
  either humans or machines
• Defining knowledge and baseline taxonomy of an
  organization
• The scope and level questionnaires and
  interview techniques
• Entire company or a specific organization
• Determine detail level (granularity) of knowledge
  analysis
• Trade-off of granularity
• Analyzing the source of knowledge
• Operation manuals, meeting minutes, external
  data, project deliverables, customer contact
  records

It is important to determine the proper level of
detail to meet organizational knowledge demand
16
4. Procedures of buildingthe knowledge map(3/10)

• Step.2 Process map analysis
• Extract organizational knowledge base (experience
  and know-how) on the business process
• Process map
• Process, Flow (dependency), Event, External object

ltAn example of a process mapgt
17
4. Procedures of buildingthe knowledge map(4/10)

• Step.3 Knowledge extraction
• Three types of extracted knowledge
• Prerequisite knowledge before process execution
• Used knowledge during execution
• Produced knowledge after execution
• Knowledge extraction techniques
• Interviewing (structured, unstructured,
  semi-structured) domain experts using a prepared
  questionnaire
• Document analysis documents, such as operation
  manual, organization chart, training material and
  external documents.
• System analysis information stored in database,
  including system log and file structure
• Knowledge workshop (KW) in a specified
  knowledge area. KW is highly interactive.
• ontology definition, building knowledge consensus
  among participants, knowledge extraction session

18
4. Procedures of buildingthe knowledge map(5/10)

• Step.4 Knowledge profiling
• Describe knowledge with pre-defined items
  (attributes) and derive relationships with
  process
• Junnarkar (1997)
• knowledge management has two aspects connecting
  people with information and connecting people
  with people
• Informational attributes
• keywords, description, importance, and
  people-finder

19
4. Procedures of buildingthe knowledge map(6/10)
ltSample attributes of knowledge profilegt
20
4. Procedures of buildingthe knowledge map(7/10)

• Step.5 Knowledge linking
• Identify navigation path of knowledge.
• Each node denoting knowledge item and link
  denoting pre- and post-relationship between
  knowledge

21
4. Procedures of buildingthe knowledge map(8/10)
ltAn example of a knowledge linkgt
22
Question 2
H.B. Kim

• Q In page 20, I can see the sample knowledge
  link. and this is very simple with only two
  notation rectangles and arrowsI like this
  diagram because it's very simple and easy
  understandable. but I wonder if this diagram can
  show and explain sufficiently.
• In order to just explain a navigation path of the
  knowledge, I think the arrow having a direction
  is enough.
• And other attributes explaining knowledge
  are included in another step such as knowledge
  profiling.

23
Question 3
K.S.H

• Q In page 20,
• Authors say that make link denoting pre- and
  post-relationship between knowledge items. Then,
  in my opinion all link in page 20 should be
  bidirectional. How do you think? and, second
  simple question is what S(ij) denote in page 20?
• The direction of the arrows represents the
  direction of the knowledge. So the direction can
  be bidirectional in case of two knowledge are
  mutually related or mutually effective.
• and the S(ij) denotes the direction of the
  knowledge itself,
• i.e. S23 means the knowledge K2 is
  post-related to knowledge K3

24
Question 4
Yana

• On slide 20,
• there is a figure showes an example of a
  knowledge link. What do Kn and Sij stand for?
  Would you please explain the figure in detail?
• What are the Kn and Sij in the figure given
  in slide 20?
• Kn denotes Knowledge
• K5 a knowledge objective
• Sij denotes relations between knowledge
• S12 Knowledge 1 relates (effects on) with
  knowledge 2

25
Question 5

• I think "knowledge linking" might be one ofthe
  most difficult and important step in knowledge
  building procedure.Who usually conduct this
  "knowledge linking", I think It needs a dedicated
  expert, and,
• Isn't there any methods to do automatically
  by s/w nowadays?
• There of course, is many solutions for knowledge
  management. From standalone solution to web based
  solution. which helps constructing knowledge map
  and searching proper knowledge. And also provides
  a efficient way to building knowledge map,
• but, most companies are having their own
  characterized (distinguishing) process and work
  flows
• so, the solutions are also need some
  adoption or customization for companies own
  process

26
4. Procedures of buildingthe knowledge map(9/10)

• Step.6 Knowledge map validation
• A structured walkthrough
• domain experts, business managers, and knowledge
  map producer
• Review checkpoints
• Are all knowledge extracted? (Unidentified
  knowledge should be found)
• Is any knowledge redundant over organizations?
• Are profiles and links fully described for all
  knowledge?
• Are the knowledge profile and knowledge map
  consistent?

27
4. Procedures of buildingthe knowledge map(10/10)

• Knowledge mapping procedures deliverables
• (Speel and Shadbolt, 1999)
• Knowledge mapping deliverables a standardized
  terminology, new knowledge creation, knowledge
  maps, knowledge gaps
• Knowledge dissemination deliverables final
  report, an electronic system containing the
  captured knowledge
• People-oriented deliverables a network of
  experts

28
5. Case study P steel company(1/7)

• Overview
• P steel company
• Most competitive company with 30 year in Steel
  industry
• 15 thousand employees (regular)
• The total sales 25 billion
• Operating profit 4 billion dollars
• Net profit 3 billion dollars
• Hot rolling process is constructed

29
5. Case study P steel company(2/7)

• Step.1 Defining organizational knowledge
• Define knowledge which occurs in a hot rolling
  mill
• Knowledge workshop on ontology
• Knowledge requirements
• Analyze input sources
• Develop basic taxonomy
• Knowledge categories mechanical, electrical,
  instrumental, information system, and control

30
5. Case study P steel company(3/7)

• Step.2 Process map analysis
• Process maps based on task flow

ltProcess map of hot rolling millgt
31
5. Case study P steel company(4/7)

• Step.3 Knowledge extraction
• Interview and questionnaires (experts 20 years
  field experience)
• Five items per experts, Describe characteristics
  and evaluate

ltAn example of knowledge extractiongt
32
5. Case study P steel company(5/7)

• Step.4 Knowledge profiling
• matching the knowledge with process map based on
  interview results
• reviewing the knowledge list with domain experts
• Confirm the final knowledge profile
• title, creating date, author, expert, location,
  and a brief description

ltAn example of a knowledge profilegt
33
5. Case study P steel company(6/7)

• Step.5 Knowledge linking
• helpful in identifying knowledge flow and
  association

34
5. Case study P steel company(7/7)

• Step.6 Knowledge map validation
• A structured walkthrough with domain experts to
  validate the knowledge map

35
Recent trend

• Changing Trend
• Functional change
• Multi-dimension knowledge map multimedia etc
• Knowledge based process management
• Combined with human resource management
• Expert training, scouting
• Mileage, employee evaluations, reward
• Enterprise knowledge portal
• Provides connection between work process and
  knowledge

36
6. Conclusion(1/2)
37
6. Conclusion(2/2)

• In this paper
• Define knowledge ontology to represent
  organizational knowledge
• Analyze business process using a process map
  technique
• Extracting knowledge based on the business
  process
• Produced a detail profile of each knowledge
• Construct a knowledge map representing a
  knowledge link by investigating the pre- and
  post-relationship
• Lessons

Maintaining knowledge is more difficult than
creating knowledge. Promotion of knowledge
sharing culture is indispensable. Top management
support on knowledge management project is
Inevitable. Reward system is clearly declared to
enhance knowledge sharing. Knowledge management
system should satisfy knowledge requirement.
38
Thank You !
Building the knowledge map