Transcript Enginyeria del Software I - Research Group on Artificial Intelligence

Knowledge Management
D. Riaño
Knowledge Management
1

Índex
•
•
•
•
•
•
•
Introduction
History
Knowledge Model and Knowledge Life Cycle
Representation
KM technologies
KM tools
Specific purpose technologies
D. Riaño
Knowledge Management
2
INTRODUCTION
D. Riaño
Knowledge Management
3
Definition
Knowledge management (KM) is ...
• ... the process through which organizations generate value from
their intellectual property and knowledge-based assets. KM involves
the creation, dissemination, and utilisation of knowledge.
• ... the strategy, processes, and technology employed to enable an
enterprise to acquire, create, organise, share, and make actionable
knowledge needed to achieve the vision of the enterprise.
• ... the tools, techniques, and strategies to retain, analyse, organise,
improve, and share business expertise.
WAY to perform some TASK aiming to some GOAL through knowledge
D. Riaño
Knowledge Management
4
Melt of disciplines
Artificial
Intelligence
Management
Science
KM
Information
Retrieval
Organizational
Behaviour
D. Riaño
Knowledge Management
5
Four Disciplines
• Management Sciences:
Management sciences are a range of methods used to assist managers through
applying scientific and quantitative approaches to the management of
organizations, often involving the construction of computable models of the key
features in decision-making.
• Organizational Behaviour:
Organizational Behaviour is the study of human behaviour at the individual, group
and organizational level.
• Artificial Intelligence:
Artificial Intelligence is a branch of science which deals with helping machines find
solutions to complex problems in a more human-like fashion. This generally
involves borrowing characteristics from human intelligence, and applying them as
algorithms in a computer friendly way.
• Information Retrieval:
Information retrieval is the task of finding information.
D. Riaño
Knowledge Management
6
Knowledge in Business
OLD VISION
•
•
•
•
•
NEW VISION
Francis Bacon’s vision
Knowledge is power
Foster individualism & competition
Company output: products
Modernization trough new
technologies.
•
•
•
•
•
KM’s vision
Sharing knowledge is power
Foster grouping & collaboration
Company outputs: services and
products derived from knowledge
Modernization through incorporating
knowledge at decisional level.
“Knowledge itself is worthy of attention because it tells firms how to do things and
how they might do them better”
T. H. Davenport, Director of the Accenture Institute for Strategic Change
L. Prusak, Executive Director of the IBM Institute for KM
D. Riaño
Knowledge Management
7
Knowledge Worldwide
People leaving a firm.
Lost of Knowledge.
Capturing company knowledge.
KM tools.
D. Riaño
Scientific & Technical migration.
Scientist going back to their
born countries.
Knowledge migration from rich
to new emerging countries.
Capturing K in rich countries.
KM Policies.
Knowledge Management
8
General Objectives of KM
1. Strategic management of the
intellectual resources.
2. Efficient K discovery.
3. Effective K application:
•
Business Strategies
Products and Services
Business Processes
Organisational Structures
Policies and Procedures
Culture and Values
Information Systems
Utilisation of the Available K
Knowledge Sharing and Reuse
Accessibility of Knowledge
Embedding K in the Work Context
Knowledge processes:
1. production
2. validation
3. integration
D. Riaño
Aspects of the enterprise that KM
deals with:
•
The enterprise perspective:
What’s what a company knows?
How efficiently it uses what it knows?
How it acquires and uses new K?
Knowledge Management
9
KM Model: Software Experience Factory
Experience Factory
Knowledge
Base
Organization
D. Riaño
•Structure
•Resources
•Norms
•Strategies
KM
Tool
Project
Knowledge Management
team
infrastructure
work plan
budget
Decisions
&
Evaluations
10
Data, Information, Knowledge, Wisdom, …
Life Cycles
WISDOM
Engineering
KNOWLEDGE
Engineering
INFORMATION
D. Riaño
Reuse
KM
Represent
Engineering
Acquire
DATA
Engineering
•Connectivity
• Transactions
• Informativeness
• Usefulness
• Cost
• Speed
• Capacity
• Timeliness
• Relevance
• Clarity
Quantitative
Evaluation
Knowledge Management
Qualitative
Evaluation
11
Data
• A) set of discrete, objective facts about events. Data is
transformed into information by adding value through
context, categorisation, calculations, corrections, and
condensation.
• B) facts and figures, without context and interpretation.
• The nature of data is raw and without context. It simply
exists and has no significance beyond its existence. It
can exist in any form, usable or not.
Single value: 90 kg.
Multiple value: (green, ugly, biped, grumpy)
D. Riaño
Knowledge Management
12
Information
• A) a message, usually in the form of a document or an
audible or visible communication meant to change the
way the receiver perceives something, to have an
impact on his judgement and behaviour.
• B) patterns in the data.
• Information is data that have been given a meaning by
way of context.
Single value: 90 kg.
Multiple value: (green, ugly, biped, grumpy)
D. Riaño
Knowledge Management
13
Knowledge
•
•
•
•
•
A fluid mix of framed experience, values, contextual information, and expert insight
that provides a framework for evaluating and incorporating new experiences and
information. It originates and is applied in the minds of “knowers”. In organisations,
it often becomes embedded not only in documents or repositories but also in
organisational routines, processes, practices, and norms.
Actionable information.
The integration of ideas, experience, intuition, skill, and lessons learned that has
the potential to create value for a business, its employees, products and services,
customers and ultimately shareholders by informing decisions and improving
actions.
Knowledge is information combined with understanding and capability; it “lives” in
the minds of people. Typically, knowledge provides a level of predictability that
usually stems from the recognition of patterns.
Knowledge is information that has been generalized to increase applicability.
Hero + Fun + Reward = successful road movie
D. Riaño
Knowledge Management
14
Data + “meaning” = Information
• Sorts of “meanings”:
– Contextualization: the purpose of the data gives a meaning.
(ex. Clients that will be emailed)
– Categorization: the data are classified / generalized in concepts.
(ex. Company clients vs. Autonomous clients)
– Calculation: the meaning is given by a mathematical or statistical
analysis.
(ex. Good client = buys  1$ million)
– Correction: remove errors from data.
(ex. Expenses in £ (instead of €) inform about English clients)
– Condensation: data is summarized in a more concise form.
(ex. Incentives out of client data gives info about incentive plans)
D. Riaño
Knowledge Management
15
Information + “something” = Knowledge
• Sorts of “something”:
– Comparison: is this information representing something similar to
other situations.
(ex. Defining a firm crisis)
– Consequences: implications of the information in company
decisions and actions.
(ex. Identify moments in which the firm must invest)
– Connections: how the information is related to other information.
(ex. There is a ratio 2/1 between incomes and investment)
– Conversations: what people think about some information.
(ex. Useful / useless concepts)
• Something = application (Tobin, 1998)
D. Riaño
Knowledge Management
16
Knowledge + intuition + experience = Wisdom
• Other upper to Knowledge concepts:
– Wisdom:
knowledge + intuition + experience
– Expertise:
wisdom + selection + principles + constrains + learning
– Capability:
expertise + integration + distribution + navigation
D. Riaño
Knowledge Management
17
Sorts of Knowledge (i): evidence
• Explicit Knowledge: the kind of knowledge which can be expressed
in words and numbers and shared in the form of data, scientific
formulae, product specifications, manuals, universal principles, etc.
This kind of knowledge can be transmitted across individuals
formally and systematically. It can be processed by a computer,
transmitted electronically, or stored in databases.
• Implicit or Tacit Knowledge: the kind of knowledge which can be
found in the heads of employees, the experience of customers and
the memories of past vendors. It is highly experiential, difficult to
document in any detail, ephemeral and transitory.
TO
FROM
D. Riaño
Tacit
Explicit
Tacit
Socialization
Internalization
Explicit
Externalization
Combination
Knowledge Management
18
Sorts of Knowledge (ii): purpose
• Declarative Knowledge or know-what: factual assertions
an organisation makes about itself, its capabilities, and
the marketplace. With this knowledge you know what are
the tasks that you have to do.
• Procedural Knowledge or know-how: business and
organisational processes and strategies of the company.
With this knowledge you know how you are supposed to
do the tasks that you have to do.
We do what we do because of of our know-what
We do what we do the way we do it because of our know-how
D. Riaño
Knowledge Management
19
Sorts of Knowledge (iii): ownership
• Individual Knowledge: personal skills, expertise, and
experience of each employee of a company about the
company processes and the company related domains.
• Group Knowledge: understanding of company groups of
employees (i.e. collectives) as they collaborate and cooperate. This includes all the individual knowledge of
each of the employees in the group and some extra
added value.
• Organizational Knowledge: knowledge held by the
organization as a whole.
D. Riaño
Knowledge Management
20
Sorts of Knowledge (iv): format
• Informal Knowledge: natural language oral, textual or
graphical representation of the knowledge (ex. *.TXT).
• Semi-Structured Knowledge: informal representation of
knowledge enriched with some attributes (ex. *.XML).
• Structured Knowledge: the knowledge is represented
according to some attribute-based structures (ex. *.DB2)
• Formal Knowledge: the knowledge is represented by
means of knowledge structures as frames, production
rules, ontologies, etc.
D. Riaño
Knowledge Management
21
Crossing Enterprise Aspects and Knowledge Types
Business Strategies
Products and Services
Business Processes
Organisational Structures
Policies and Procedures
Culture and Values
Information Systems
D. Riaño
explicit
implicit
Y
Y
N
Y
Y
Y
Y
N
N
Y
N
Y
Y
Y
Knowledge Management
know-what know-how
N
Y
N
Y
Y
Y
Y
Y
N
Y
N
Y
N
N
22
Data representation and organization

• Matrix representations
– Column Heading = typed feature
– Row = instance
– Cell = (single) data
arity
data
• Data bases
– Relationship: column to column
– Cardinality: 1, N
– Optionality: 0 allowed Y/N
• Data warehouses
D. Riaño
Knowledge Management
23
Information representation

• Information Systems
D. Riaño
Knowledge Management
24
Knowledge representation and modelling
KR aims at expressing knowledge in a computer manageable
way, so that it can be used in an computer intelligence process.
•
KR aspects:
–Syntactic: structures that support the representation.
–Semantic: meaning of the knowledge represented.
–Reasoning & Inference: process by which knowledge is used to obtain conclusions.
•
Inference aspects:

–Forward chaining (modus ponens): A, A B
–Backward chaining (modus tollens): ¬B, A B

•
B
¬A
Knowledge-base aspects:
–Completeness: given a KB, the inference process can find B or ¬B, for any correct
assertion B.
–Soundness: given a KB, the inference process cannot find both B and ¬B, for any
correct assertion B.
D. Riaño
Knowledge Management
25
Artificial Intelligence Knowledge Models
•
•
•
•
•
Frames (Minsky 1975)
Scripts
Semantic Networks (Michalski 1983)
Rules
Ontologies
• Tools to model knowledge:
–
–
–
–
D. Riaño
commonKADS
Protégé 2000
Unified Modelling Language (UML) - Object Constraint Language (OCL)
Multi-Perspective Modelling
Knowledge Management
26
Knowledge Representation: Frames
•
•
•
•
•
Frame: a single know-what knowledge structure containing slots.
Slot: element of the frame that contains one or more facets.
Facets: element that describes something about a slot.
Demons: procedures attached to slots that are fired circumstantially.
Instance: frame example.
•
Relationships between frames:
–
–
–
•
Slot sub-concepts: contains links to other frames which represent sub-concepts.
Slot type: GENERIC or INSTANCE.
Slot with facet other containing another frame.
Facets may take one of the following forms:
–
–
–
–
–
–
D. Riaño
Values: contains the slot (single or multiple) value.
Default: used if there is not other value present.
Range: informs about the kind of information the slot can contain.
if-added: procedural attachment which specifies an action to be taken when a value in the
slot is added or modified (forward chaining, data-driven, event-driven or bottom-up
reasoning).
if-needed: procedural attachment which triggers a procedure which goes out to get
information which the slot doesn't have (backward chaining, goal driven, expectation driven
or top-down reasoning).
Other: may contain frames, rules, semantic networks, or other types of knowledge.
Knowledge Management
27
Frames: car’s example
(frame
(name
(type
(sub-concepts
(company
(model
(horse-power
(start-prod
(finish-prod
(color
(factory-price
(retail-price
)
(wheels
(values CLASSIC-CAR))
(values GENERIC))
(range BEETLE SEDAN JEEP TOPOLINO …))
(range CAR-COMPANY)
(if-needed (search-Co model)))
(range CAR-MODEL)
(if-added (confirm-exists model company)))
(range 1..200))
(default UNKNOWN))
(default PRESENT))
(range {R W B Y DARK OTHER})
(range NUMBER))
(if-needed
(add-interests factory-price))
(if-added
(check-above-15% factory-price)))
(range NUMBER) (default 4))
(frame
(name
(values BEETLE))
(type
(values GENERIC))
(instances (values John’s-CAR …))
(company (values VOLKSWAGEN))
(Horse-Power
(range 50..90))
(start_prod
1938)
(color
B)
(retail-price
8000€)
)
D. Riaño
Knowledge Management
28
Knowledge Representation: Scripts
• Script: A structure that describes appropriate sequences of events
in a particular context. A type of frame that describes what happens
temporally (know-how).
• Properties: objects being part of the script (frames or strings).
• Roles: agents involved in the script definition (frames or strings).
• Starting/Opening conditions: conditions that make the script be
valid (pre-condition).
• Scenes: actions in the script.
• Results: conditions that are valid after the script is ran (postcondition).
• Scripts extend frames with complex temporal events.
D. Riaño
Knowledge Management
29
Scripts: car’s example
(script
(name
(type
(props
(roles
(opening
(results
(scenes
(values BUY-A-CAR))
(values GENERIC))
(values SHOP MONEY CAR CATALOG OFFICE))
(values CUSTOMER SELLER))
(wants CUSTOMER CAR))
(if-needed
(owner CAR CUSTOMER)
(has-less-money CUSTOMER)
(increase-sells SHOP)))
(ENTERING
(enters CUSTOMER SHOP)
(go-to-scene INSPECTING))
(INSPECTING
(observes CUSTOMER CAR)
(or (go-to-scene ASKING) (leaves SHOP CUSTOMER)))
(ASKING
(look-for CUSTOMER SELLER)
(meet CUSTOMER SELLER OFFICE)
(asks-for CUSTOMER CATALOG)
(informs SELLER CUSTOMER)
(or (go-to-scene BUYING) (leaves SHOP CUSTOMER)))
(BUYING
(pays CUSTOMER MONEY SELLER)
(leaves SHOP CUSTOMER))
))
D. Riaño
Knowledge Management
30
Knowledge Representation: Semantic Networks
(John F. Sowa)
•
•
•
•
•
•
Definitional networks emphasize the is-a relation between concepts. The resulting
network, also called a generalization or subsumption hierarchy, supports the rule of
inheritance for copying properties defined for a supertype to all of its subtypes. Since
definitions are true by definition, the information in these networks is often assumed
to be necessarily true.
Assertional networks are designed to assert propositions. Unlike definitional
networks, the information in an assertional network is assumed to be contingently
true, unless it is explicitly marked with a modal operator. Some assertional networks
have been proposed as models of the conceptual structures underlying natural
language semantics.
Implicational networks use implication as the primary relation for connecting nodes.
They may be used to represent patterns of beliefs, causality, or inferences.
Executable networks include some mechanism, such as marker passing or
attached procedures, which can perform inferences, pass messages, or search for
patterns and associations.
Learning networks build or extend their representations by acquiring knowledge
from examples. The new knowledge may change the old network by adding and
deleting nodes and arcs or by modifying numerical values, called weights, associated
with the nodes and arcs.
Hybrid networks combine two or more of the previous techniques, either in a single
network or in separate, but closely interacting networks.
D. Riaño
Knowledge Management
31
Knowledge Representation: “Definitional” Semantic Networks
D. Riaño
Knowledge Management
32
Knowledge Representation: “Assertional” Semantic Networks
A
b
C
A C: b(A,C)
b
C
A C: ¬b(A,C)
¬
A
Example: “If a person wants a car, he must go to the car dealer”
¬
person
want
¬
D. Riaño
go
car
dealer
Knowledge Management
33
Knowledge Representation: “Implicational” Semantic Networks
• Semantic network in which arcs represent logic implications.
• Sorts of “implicational” Semantic Networks:
–
–
–
–
Belief Networks (Judea Pearl, 1988)
Causal Networks (Chuck Riegel 1976)
Bayesian Networks
Truth-Maintenance Systems, TMS (Doyle, 1979)
Example: “A person goes to a car dealer because he needs a car,
and buy it if he likes the car and he can pay the price”
Need car
Go dealer
good
Like car
deal
Can pay
bad
Go home
Buy car
D. Riaño
Knowledge Management
34
Knowledge Representation: “Executable” Semantic Networks
• Semantic networks that represent dynamic processes or procedural
knowledge.
• General elements of the networks:
– Message passing through the network arcs
– Attached procedures to the network nodes
– Graph transformations as external triggered actions
• Sorts of “executable” Semantic Networks:
– Dataflow diagrams
– Petri Nets (Carl Adam Petri, 1962)
D. Riaño
Knowledge Management
35
“Executable” Semantic Network: examples
DFD: “Retail price calculation”
Petri Net: “Car selling”
Factory price
Company
%
Company
price
entering
car inspecting
Shop
%
Shop
Profit margin
Dealer
%
waiting room
going
available dealer
+
Factory price
Dealer
Profit margin
Retail price
asking
1 Calculate
Company
profit
Factory price
Retail price
SHOP PROFIT
2 Calculate
Shop
profit
D. Riaño
buying
3 Calculate
Dealer
profit
4 Calculate
Car Price
Knowledge Management
36
Knowledge Representation: “Learning” Semantic Networks
• Semantic networks that can adapt to new incoming
evidences.
• These modifications can be at three levels:
– Rote memory: the new knowledge is represented by a semantic
network that is appended to the global semantic network.
– Changing weights: when the knowledge in the network is
weighted with numerical values (in nodes and arcs), the new
knowledge modifies some of the weights in the network.
– Restructuring: new knowledge changes the structure of the
semantic network adding or removing nodes and arcs.
• Sorts of “learning” Semantic Networks:
– Artificial Neural Networks
D. Riaño
Knowledge Management
37
Knowledge Representation: Rules
• Selectors
• Premise  Conclusion
• Syntactic differentiation
–
–
–
–
–
Conjunctive:
Disjunctive:
K-term DNF:
K-DNF:
K-CNF:
[a1 a2  …  ak  b].
[a1  a2  …  ak  b].
[(a11  …  a1k1)  …  (ai1  …  aiki)  b], i k.
[(a11  …  a1k1)  …  (ai1  …  aiki)  b], kj k.
[(a11  …  a1k1)  …  (ai1  …  aiki)  b], kj k.
• Semantic differentiation
– Production rules: conceptual rules.
– Association rules: the rule indicates the value of b, when the values of
the a’s are known.
– M-of-N rules: the rule is fired if M of N selectors in the premise are true.
– Ripple-down rules: exceptions to the rules are appended at the end of
the rule as a ripple down rule.
D. Riaño
Knowledge Management
38

Knowledge Representation: Ontologies
•
•
•
An ontology is a specification of a conceptualization.
An ontology may take a variety of forms, but necessarily it will include a
vocabulary of terms, and some specification of their meaning. This includes
definitions and an indication of how concepts are inter-related which
collectively impose a structure on the domain and constrain the possible
interpretations of terms.
What does an ontology do?
–
–
–
–
•
Captures knowledge
Creates a shared understanding – between humans and for computers
Makes knowledge machine processable
Makes meaning explicit – by definition and context
Components of an ontology:
–
–
–
–
–
Concepts: Class of individuals
Relationships between concepts
Is a kind of relationships: they form a taxonomy
Other relationships: they give further structure –is a part of, belongs to, etc.
Axioms: constrains about the concepts –Disjointness, covering, equivalence, etc.
Ex. Cover (X, Y) <- X member Of interval and Y member Of interval and X.start <= Y.start and X.end >= Y.end
•
Instances
D. Riaño
Knowledge Management
39

strong semantics
Modal Logic
First Order Logic
Logical Theory
Is Disjoint Subclass of
Description Logic
with transitivity
DAML+OIL, OWL
property
UML
Conceptual Model
RDF/S
XTM
Extended ER
Thesaurus
ER
Is Subclass of
Has Narrower Meaning Than
DB Schemas, XML Schema
Taxonomy
Semantic Interoperability
Structural Interoperability
Is Sub-Classification of
Relational
Model, XML
Syntactic Interoperability
weak semantics
D. Riaño
Knowledge Management
40
Knowledge Engineering
• Formal methodologies for developing knowledge-based
systems.
• KB and KB systems: Expert Systems.
• K Life Cycle: problem selection, knowledge acquisition,
knowledge
representation,
knowledge
encoding,
knowledge testing and evaluation, implementation and
maintenance.
CREATE
SELECT
ACQUIRE
KM
APPLY
D. Riaño
SHARE
REPRESENT
KE
TEST & EVALUATE
Knowledge Management
ENCODE
41
Knowledge Acquisition
Knowledge Expert
Knowledge Engineer
Domain overview, goals, etc.
Identified concepts, values, etc.
Identified sources of information
Knowledge
Validation
Identified relationships, sequences, etc.
Knowledge
Verification
Amendments
Knowledge representation
D. Riaño
Knowledge Management
42
The KM process
•
•
•
•
•
•
•
•
Determine goals for KM activities
Create an overview of the available knowledge
Structure and Integrate knowledge
Acquire knowledge
Goal oriented disseminate the knowledge
Use productively the knowledge for the company
Storage and Maintain the knowledge
Assess the current knowledge and the compliance with goals
D. Riaño
Knowledge Management
43
Case-Based Reasoning
A branch of AI that attempts to combine the power of narrative with
the codification of knowledge on computers. Involves extraction of
knowledge from a series of narratives, or cases, about the problem.
(Aamodt & Plaza, 1984) The CBR paradigm covers a range
of different methods for organizing, retrieving, utilizing
and indexing the knowledge retained in past cases.
Cases may be kept as concrete experiences, or as a
generalization of a set of similar cases. Cases may be
stored as separate knowledge units and distributed within
the knowledge structure. Cases may be indexed by a
prefixed or open vocabulary, and within a flat or hierarchical
index structure. The solution from a previous case may be
directly applied to the present problem, or modified
according to differences between the two cases. The
matching of cases, adaptation of solutions, and learning
from an experience may be guided and supported by a
deep model of general domain knowledge, by more shallow
and compiled knowledge, or be based on an apparent,
syntactic similarity only.
D. Riaño
Knowledge Management
44
Knowledge-Based Systems: Expert Systems
First generation ES
KNOWLEDGE
EXPERT
PROBLEM
SELECTION
Second generation ES
KNOWLEDGE
ENGINEER
KBS
KNOWLEDGE
ACQUISITION
DB
KNOWLEDGE
USER
KNOWLEDGE TESTING
AND EVALUATION
KNOWLEDGE CODIFICATION
KNOWLEDGE REPRESENTATION
KBS
KNOWLEDGE BASE
KNOWLEDGE BASE
EXPLAINATION
FACILITY
INFERENCE
ENGINE
SEARCH STRATEGIES
KNOWLEDGE
USER
USER INTERFACE
D. Riaño
MACHINE LEARNING
Knowledge Management
EXPLAINATION
FACILITY
INFERENCE
ENGINE
SEARCH STRATEGIES
USER INTERFACE
45
Knowledge (Management) Systems
•
•
•
Specialised systems that interact with the organisation’s systems to facilitate
all aspects of knowledge processing.
They have evolved from Knowledge-Based Systems.
Unlike KB systems, KM systems must fulfil the following requirements:
–
–
–
–
–
–
Supply a conceptual level
Reuse the existing K
Convenient and save adaptation to individual needs
Intuitive understanding
Support of multiple perspectives
Integration of perspectives
The MEMO Architecture (Frank 97)
“Multi-perspective Enterprise MOdeling”
D. Riaño
Knowledge Management
46
Knowledge Structures: KM vision
Sustaining & Extending
a K-Sharing Culture
KM – Knowledge Management
CoP – Communities of Practice
Full
Implementation
KM Pilots &
Measurement
KM Tools &
Technologies
KM
Awareness
KM
Strategy
Change
Management
KM
CoP Building
Organization & Nurturing
KM Target
KM
KM
Areas
Taxonomy Benchmark
“Knowledge Management – Learning from Knowledge Engineering” – Jay Liebowitz
D. Riaño
Knowledge Management
47
Knowledge Structures: Semantic Web vision
“The Semantic Web will globalize KR, just as the WWW globalize hypertext” -- Tim Berners-Lee
D. Riaño
Knowledge Management
48
PERSPECTIVE
D. Riaño
Knowledge Management
49
(Pre-)Historical Evolution of KM
• Oral Knowledge Transmission
– Story-teller
– Mankind traditions
• Textual and Graphical Knowledge Transmission
– Documents and File Cabinets
– Books
• Computer-Based Knowledge Transmission
–
–
–
–
–
D. Riaño
Email
Intranets and Internet
Magnetic, laser-based ,etc. file record systems
Information Systems
Knowledge Bases
Knowledge Management
50
History of Computer-Based KM
• FGKM: First Generation Knowledge Management
– focused on the use of technologies to help users to extract knowledge
and share this knowledge within the enterprise.
– vision: valuable knowledge already exists.
– tools: technology always seems to provide the answer.
– purposes:
• enhance the deployment of knowledge into practice.
• knowledge integration.
• SGKM: Second Generation KM or “the new KM”
– focused on the use of technologies to generate new valuable
knowledge, validate this knowledge and integrate it in the enterprise
business processes and business strategies.
– vision: knowledge is something that is produced.
– purposes:
• knowledge production and integration.
• Third Generation KM
D. Riaño
Knowledge Management
51
First Generation Knowledge Management
•
•
•
•
•
•
•
Groupware
Information Indexing and Retrieval Systems
Knowledge Repositories
Data Warehousing
Document Management
Imaging
Data Mining
D. Riaño
Knowledge Management
52
Second Generation Knowledge Management
•
•
•
•
•
•
•
•
•
•
Supply-Side vs. Demand-Side KM
The Knowledge Life Cycle
Knowledge Processes
Knowledge Rules
Knowledge Structures
Nested Knowledge Domains
Organizational Learning
The Open Enterprise
Complexity Theory
Sustainable Innovation
D. Riaño
Knowledge Management
53
FGKM: Groupware or “electronic collaboration”
“Software that supports the ability for two or more people to
communicate and collaborate”
(P. & T. Johnson-Lenz, 1978)
ALTERNATIVE TECHNOLOGIES
1.
2.
3.
4.
5.
6.
7.
8.
9.
10.
11.
12.
D. Riaño
Email and messaging
Group calendaring and scheduling
Electronic Meeting Systems (EMS)
Desktop video and real-time data conferencing (synchronous)
Non real-time data conferencing (asynchronous)
Group document handling
Workflow
Group utilities and development tools
Groupware services
Groupware and KM frameworks
Groupware applications
Collaborative Internet-based applications and products
Knowledge Management
54
FGKM: Information Indexing and Retrieval
• Information = (data, meaning)
• meaning: unique or not
• Information systems:
–
–
–
–
Disordered lists: slow access, impractical.
Ordered lists as Yellow & White pages: dichotomy fast access.
Hierarchical indices: fast access.
Hash tables: instantaneous access.
1. Data1
2. Data2
…
N. DataN
D. Riaño
H(m)=
meaning
Data1, Data2, …, DataN
Knowledge Management
meaning
Data1, Data2, …, DataN
55
FGKM: Knowledge Repositories

• Organisational Memory or Knowledge Repository: computer system
that continuously captures and analyses the knowledge assets of an
organisation. It is a collaborative system where people can query
and browse both structured and unstructured information in order to
retrieve and preserve organisational knowledge assets and facilitate
collaborative working.
• Knowledge-base: case-based, ontology-based, …
• Types (Davenport & Prusak 1998):
– External knowledge (e.g. competitive or business intelligence: selection,
collection, interpretation and distribution of publicly-held information that
has strategic importance)
– Structured Internal knowledge (e.g. reports & documents)
– Informal Internal knowledge (e.g. discussion databases)
• Models
– Knowledge network model: person-to-person
– Knowledge repository model: person-to-repository-to-person.
– Hybrid: combination of both.
D. Riaño
Knowledge Management
56
FGKM: Data Warehousing
•
•
•

A data warehouse is a copy of transaction data specifically structured for querying
and reporting.
Se llama DataWarehouse al almacén de datos que reúne la información histórica
generada por todos los distintos departamentos de una organización, orientada a
consultas complejas y de alto rendimiento. Un DataWarehouse pretende conseguir
que cualquier departamento pueda acceder a la información de cualquiera de los
otros mediante un único medio, así como obligar a que los mismos términos tengan
el mismo significado para todos. Un Datamart es un almacén de datos históricos
relativos a un departamento de una organización, así que puede ser simplemente
una copia de parte de un DataWarehouse para uso departamental.
Tanto el DataWarehouse como el Datamart son sistemas orientados a la consulta, en
los que se producen procesos batch de carga de datos (altas) con una frecuencia
baja y conocida. Ambos son consultados mediante herramientas OLAP (On Line
Analytical Processing) que ofrecen una visión multidimensional de la información.
Sobre estas bases de datos se pueden construir EIS (Executive Information
Systems, Sistemas de Información para Directivos) y DSS (Decision Support
Systems, Sistemas de Ayuda a la toma de Decisiones). Por otra parte, se conoce
como Data Mining al proceso no trivial de análisis de grandes cantidades de datos
con el objetivo de extraer información útil, por ejemplo para realizar clasificaciones o
predicciones.
D. Riaño
Knowledge Management
57
FGKM: Document Management
•
•
Document management is the process of managing documents through their lifecycle. From
inception through creation, review, storage and dissemination all the way to their destruction.
The result of a document management system will be an immediate access to information
benefiting companies, their partners and their customers:
–
–
•
Shortened time frames to produce information requested.
Better decisions enabled by accurate, timely and accessible information will improve the quality of work.
Document management involves:
–
–
–
–
–
•

Authors that create documents, add content, and refine it.
Editors that oversee the documents to ensure that they have relevant content and contain useful search
terms.
Software facilities that enable authors and editors to easily and consistently manage the documents. These
facilities ensure that documents are generated to current digital library standards and so enables better
resource discovery.
Publishing is the process of accepting the authors work, assisting to refine the content, and making the
document publicly available.
Promotion is the process of expose the documents. It involves ensuring that the catalogue itself is wellknown and that the documents can be discovered through many avenues.
Example: The Standard Generalized Markup Language (SGML) is an international standard for
the definition of device-independent, system-independent methods of representing text in
electronic form. SGML is a meta language, that is, a means of formally describing a language. The
Document Type Definition (DTD) defines the metadata elements, and their order, structure, and
relationships in the SGML document management solution. The eXtensible Markup Language
(XML) defines well-structured documents that conform to a set of rules established by a DTD.
D. Riaño
Knowledge Management
58
FGKM: The Dublin Core
•
•
The Dublin Core is a set of predefined properties for describing documents.
The first DC properties were defined in Dublin (Ohio) in 1995 and is
currently maintained by the Dublin Core Metadata Initiative.
Property
Definition
Contributor
An entity responsible for making contributions to the content of the resource
Coverage
The extent or scope of the content of the resource
Creator
An entity primarily responsible for making the content of the resource
Format
The physical or digital manifestation of the resource
Date
A date of an event in the lifecycle of the resource
Description
An account of the content of the resource
Identifier
An unambiguous reference to the resource within a given context
Language
A language of the intellectual content of the resource
Publisher
An entity responsible for making the resource available
Relation
A reference to a related resource
Rights
Information about rights held in and over the resource
Source
A Reference to a resource from which the present resource is derived
Subject
A topic of the content of the resource
Title
A name given to the resource
TypeD. Riaño
The nature or genre of theKnowledge
content of the Management
resource
59

FGKM: Imaging
• The capture and storage of electronic information from
hard-copy documents.
D. Riaño
Knowledge Management
60
FGKM: KDD & Data Mining
D. Riaño
Knowledge Management
61
SGKM: Supply-Side versus Demand-Side
•
FGKM: Supply-Side KM
(1) “It’s all about capturing, codifying, and sharing valuable knowledge”.
(2) “It’s all about getting the right information to the right people at the right
time”.
(3) “If we only knew what we know”
(4) “Knowledge is something that is there”
(5) “We need to capture and codify our tacit and explicit knowledge before
it walks out the door”
(6) “The purpose of KM is to enhance the deployment of K into practice”
•
SGKM: Demand-Side KM
(1) “It’s all about contributing to the knowledge life cycle”
(2) “Knowledge is something that we produce in human social systems,
though individual and shared processes”
(3) “The purpose of KM is to enhance knowledge production”
D. Riaño
Knowledge Management
62
SGKM: The Knowledge Life Cycle
D. Riaño
Knowledge Management
63
“Simplified” Knowledge Life Cycle
Knowledge
Production
D. Riaño
Knowledge
Claims
Knowledge
Validation
Knowledge Management
Organisational
Knowledge
Knowledge
Integration
64
Alternative KM life cycles (Liebowitz, 2000)
1.
2.
3.
4.
5.
6.
7.
8.
9.
Transform Information into Kwlg.
Identify & Verify Knowledge
Capture & Secure Knowledge
Organize Knowledge
Retrieve & Apply Knowledge
Combine Knowledge
Create Knowledge
Learn Knowledge
Distribute/Sell Knowledge
D. Riaño
Knowledge Management
65
Alternative KM life cycles (Liebowitz&Beckman, 2000)
STAGE 1: IDENTIFY
STAGE 2: CAPTURE
STAGE 3: SELECT
STAGE 4: STORE
STAGE 5: SHARE
STAGE 6: APPLY
STAGE 7: CREATE
STAGE 8: SELL
D. Riaño
Identify is to determine competencies, sourcing strategy,
and knowledge domains.
Capture the existing knowledge is formalized during this
phase.
Select consists on assessing knowledge relevance,
value and accuracy, and resolve conflicting
knowledge.
Store: The knowledge is stored by representing the
corporate memory in a knowledge repository with
various knowledge schema.
Share: Then, the stored knowledge can be shared and
finally applied in making decisions, solving
problems, automating or supporting work, job aids,
and training.
Create: New knowledge can be discovered (with or
without the use of the previous one) through
research, experimenting, and creative thinking.
Sell: Apart of applying the knowledge in stage 6, it can
be also sell. That’s to say, new knowledge-based
products and services can be developed and
marketed.
Knowledge Management
66
Alternative KM life cycles (Marquardt, 1996), …
(Marquardt, 1996)
1.
2.
3.
4.
Acquisition
Creation
Transfer and utilization
Storage
1.
2.
3.
4.
(Spek & Spijkervet, 1997)
Developing new knowledge
Securing new & existing K.
Distributing Knowledge
Combining available K.
1.
2.
3.
1.
2.
3.
4.
5.
6.
1.
2.
3.
4.
(O’Dell, 1996)
1.
2.
3.
4.
5.
6.
(Ruggles, 1997)
Generation:Creation, Acquisition, Synthsis, Fusion, Adaptation 7.
Codification: Capture, Representation
Transfer
1.
2.
(Holsapple & Joshi, 1997)
3.
Acquiring Knowledge: Extracting, Interpreting, Transferring
4.
Selecting Knowledge: Locating, Retrieving, Transferring
5.
Internalizing Knowledge: Assessing, Targeting, Depositing
6.
Using Knowledge
7.
Generating Knowledge: Monitoring, Evaluating, Producing,
Transferring
Externalizing Knowledge: Targeting, Producing, Transferring
1.
D. Riaño
(Wiig, 1993)
Creation and Sourcing
Compilation and Transformation
Dissemination
Application and Value realization
Identify
Collect
Adapt
Organize
Apply
Share
Create
(Dataware Technologies, 1998)
Identify the (business) problem
Prepare for charge
Create the KM team
Perform the knowledge audit and analysis
Define the key features of the solution
Implement the building blocks for KM
Link knowledge to people
(Van der Spek & Hoog, 1998)
Conceptualize: Make inventory of existing K., Analyze
strong and weak points
2.
Reflect: Decide on required improvements, Make plans to
improve
3.
Act: Secure, Combine, Distribute, and Develop knowledge
Knowledge Management
67 achieved
4.
Review: Compare old and new situation, Evaluate
results
The SMART KM life cycle
(Liebowitz, Rubenstein-Montaro, Buchwalter, et al. 2000)
1.
2.
3.
4.
5.
Strategize
Model
Act
Revise
Transfer
D. Riaño
- analysis
- design
- implement
- test
- implantation and update
Knowledge Management
68
The SMART KM life cycle: steps
1.1. Perform strategic planning: determine key knowledge requirements and priorities.
1.2. Perform business needs analysis: identify business problems and metrics for success.
1.3. Conduct cultural assessment and ensure knowledge sharing.
2.1. Perform conceptual modeling: conduct a knowledge audit (list types and sources of K, competencies, weaknesses, organization
and knowledge flows, etc.) and do knowledge planning (propose a KM strategy, a K. sharing culture, a cost-benefit analysis, etc.).
2.2. Perform physical modeling: develop the physical architecture.
3.1. Capture and secure knowledge: collect, verify, and evaluate knowledge.
3.2. Represent knowledge: define a formal representation “language”, classify knowledge, and encode it in the selected language.
3.3. Organize and store knowledge in the KM system.
3.4. Combine knowledge: retrieve and integrate knowledge from the entire organization.
3.5. Create knowledge: have discussion with customers and interested parties, perform exploration and discovery, conduct
experimentation, etc.
3.6. Share knowledge: distribute and make knowledge easily accessible.
3.7. Learn knowledge and go to 3.1
4.1. Pilot operational use of the KM system.
4.2. Conduct knowledge review: perform quality control (with validity, accuracy, and update metrics) and relevance review (discard
irrelevant K).
4.3. Perform KM system review: test and evaluate results.
5.1. Publish knowledge.
5.2. Coordinate KM activities and functions: activate action plans for applying knowledge and report where the knowledge is located.
5.3. Use knowledge to create value for the enterprise: sell, apply, and use the knowledge.
5.4. Monitor KM activities via metrics.
5.5. Conduct post-audit.
5.6. Expand KM activities.
5.7. Continue learning and go to back phases.
D. Riaño
Knowledge Management
69
The SMART KM life cycle: documents and products

1.1. Business needs analysis doc:
1.2. Cultural assessment and incentives doc:
2.1. Knowledge audit doc:
2.2. Visual prototype:
2.3. KM program doc:
2.4. Requirements specification doc:
3.1. Knowledge acquisition doc:
3.2. Design doc:
3.3. Visual and technical KM system prototypes:
4.1. Evaluation methodology and results doc:
4.2. KM system prototype:
4.3. User’s guide for the KM system:
5.1. Maintenance doc for KM system:
5.2. Fully functional KM system.
5.3. Post-audit doc:
5.4. Lessons learned doc:
D. Riaño
Knowledge Management
70
Knowledge Audit
Auditing knowledge for a particular target area consists on:
“identifying which knowledge is needed and available for that area,
which knowledge is missing, who has the knowledge (source), and how
it is being used (destination)”.
1. Identify the currently existing knowledge in the targeted area:
Determine existing and potential sinks, sources, flows, and constraints.
Identify and locate explicit and tacit knowledge.
Build a knowledge map of the taxonomy and flow of knowledge.
2. Identify the currently missing knowledge in the targeted area:
Perform gap analysis to detect missing knowledge.
Determine who needs the missing knowledge.
3. Provide recommendations
D. Riaño
Knowledge Management
71
Knowledge Auditing Algorithm
• What are the categories of knowledge in the targeted area?
• Which of them are currently available?
• If available,
–
–
–
–
–
–
–
–
–
–
How this knowledge is used?
What are the sources of this knowledge?
Who is using this knowledge? How often?
Who are new potential users of this knowledge?
What’s the process or processes to obtain that knowledge?
How is this knowledge adding value or benefit?
What influences this knowledge?
What are the elements that identify, use, or transform this knowledge?
How is this knowledge delivered from? Are there other delivering alternatives?
Who are the experts (in the company) in this sort of knowledge?
–
–
–
How much your work can be improved from it?
What are the potential sources of this knowledge?
What are your unanswered questions? For each one,
• What categories of knowledge do you need to do your work better?
• For all of them,
D. Riaño
•
•
•
What is the sort of knowledge missed?
Which departments/people could answer these open questions?
Which departments/people are looking for similar answers? For each one,
–
–
–
–
–
What level in the organization this department/person has?
If a person, how old is this person in the company?
Why did they ask these questions similar to yours?
Is someone in the organization putting barriers to this sort of KM?
What are the main reasons to make errors/mistakes concerning this knowledge?
Knowledge Management
72
Introducing KM practice in an enterprise: actions
1.
2.
3.
4.
5.
6.
7.
8.
9.
10.
11.
12.
13.
14.
15.
16.
(J. Liebowitz, 2000)
Obtain management buy-in.
Survey and map the knowledge landscape.
Plan the knowledge strategy.
Create/define K-related alternatives and potential initiatives.
Portray benefit expectations for KM initiatives.
Set KM priorities.
Determine key knowledge requirements.
Acquire key knowledge.
time
Create integrated knowledge transfer programs.
Transform, distribute, and apply knowledge assets.
Establish and update KM infraestructure.
Make knowledge assets.
Construct incentive programs.
Coordinate KM activities and functions enterprise-wide.
Facilitate knowledge-focused management.
Monitor Knowledge Management.
D. Riaño
Knowledge Management
73
Introducing KM practice in an enterprise: 8-step agenda
(J. Liebowitz, 2000)
1. Develop a broad vision of the KM practice and obtain
management buy-in.
2. Pursue targeted KM focus.
3. Allow team members to focus full time on KM and
build KM professional teams.
4. Install KM impact and benefit evaluation methods.
5. Implement incentives to manage knowledge.
relevance
6. Teach metaknowledge to everyone.
7. Ascertain that implemented KM activities provide
opportunities,
capabilities,
motivations,
and
permissions for individuals and the enterprise to act
intelligently.
8. Create supporting infraestructure.
D. Riaño
Knowledge Management
74
Evaluating the performance of KM in an enterprise
•
•
Knowledge work is the work produced
as a result of the use of knowledge.
Knowledge work metrics:
–
project management: measures of size,
effort, and duration of a KW project
•
•
–
•
Productivity: amount of effort required
to produce a KW project of a given size.
Delivery: time required to develop a
KW project.
quality control or defect density:
number of defects or errors in a KW
project of a given size.
Software Cost Estimation Theory
output product size

input
work hours
product size
delivery 
elapsed weeks
number of defects
defect density 
product size
productivity 
The PNR model (Putnam, Norden,
Rayleigh, 1963)
–
–
–
–
–
D. Riaño
B: skills factor
PP: productivity parameter
MBP: manpower buildup parameter
Effort: person-year units
Size o SLOC: source lines of code /K
Knowledge Management
1


size  B 3   1 

effort 

 PP   time 4 


effort
MBP 
time 3

 1
time  
 MBP



size  B

 PP

1
3




3






1
7
75
SGKM: Knowledge Processes

• Knowledge processes: any of the processes involved in
the KM life cycle.
• Knowledge processing: act of applying some knowledge
process (ex. knowledge production or integration).
• Knowledge Management is about an action that seeks to
have an impact on knowledge processing (ex. to design
a portal to enhance knowledge sharing).
D. Riaño
Knowledge Management
76
Knowledge Map
• Representation of the knowledge inside the company
• Purposes:
–
–
–
–
–
–
Generate ideas
Design a complex structure
Communicate complex ideas
Aid learning by integrating new and old knowledge
Assess understanding
Diagnose misunderstanding
• Sorts of knowledge maps:
– Organizational Maps
– Expertise Maps
– Concept Maps
D. Riaño
Knowledge Management
77
Organizational Maps
They are used to show the interactions between company members.
Albert
Manufacturing
Francine
Marketing
Eve
Guy
Bernard
Helen
Donald
Charlotte
Mary
Nora
Peter
Liz
Oscar
John
Ex: by the analysis of
the emails/internal calls
between members in
the company.
Keith
H. Resources
Management
D. Riaño
Ian
Collaborations:
Close
Distant
Isolations
Unidirectional
Hierarchies
Etc.
Knowledge Management
78
Expertise Maps
They are used to show who knows things in the company.
Area 1 AI
Dept 1
Albert
Expertise:
KM
Someone/nobody
Who/What dept.
KE
Bernard
DB
Charlotte
DSS
Donald
Eve
Marketing
Francine
Finances
Guy
Project Management
D. Riaño
Area 2
Helen
Working team
People selection
Employee formation
Etc.
Ex: by the analysis of people
participating in projects, papers,
Reports, meetings, etc. and
Their role/responsibility in that
activities.
Dept 2
Knowledge Management
79
Concept Maps
They are used to know the relationships between company concepts.
Concepts can be: objects, resources, products, etc.
AI
DM
DB
KM
DSS
KE
ExpSyst
SoftEng
D. Riaño
Knowledge Management
80

Semantic Knowledge Maps
• The links in the Map have a meaning.
• Meanings:
– Descriptive Links
C – characteristic
P – part of
T – type or subpart of
EXAMPLE
– Dynamic Links
I – influences
L – leads to
N - next
– Instructional Links
A – analogy
S – side remark
E - example
D. Riaño
Knowledge Management
81
Constructing Semantic Knowledge Maps
(Newbern & Dansereau, 1993)
1. Make a list of important concepts or main ideas.
2. For each concept or idea,
2.1. Add a node in the map, labeled with the concept.
2.2. Ask the following questions and draw links on the map,
2.2.1. Can this concept be broken down into sub concepts (T-link)?
2.2.2. For each sub concept or concept type,
2.2.2.1. What are the features of that type (C-link)?
2.2.2.2. What are the important parts of that type (P-link)?
2.2.2.3. For a each part, what are the features (C-link)?
2.2.3. What led to the starting node (L-link)?
2.2.4. What does the starting node lead to (L-link)?
2.2.5. Which things influence the starting node (I-link)?
2.2.6. What does the starting node influence (I-link)?
2.2.7. What happens next (N-link)?
2.2.8. Does anything require an analogy, remark or example (A,S,E-links)?
3. Review the map
D. Riaño
Knowledge Management
82
SGKM: Nested Knowledge Domains
• Enterprises have different levels of abstraction: the whole company,
the departments, the working groups, the individuals, etc.
• Each member of a level can have its own competencies and
therefore its own knowledge life cycle.
D. Riaño
Knowledge Management
83
SGKM: The Open Enterprise
D. Riaño
Knowledge Management

84
SGKM: Organizational Learning (OL)
• introduced by Peter Senge in 1990.
• it is “the ability to learn faster than your competitors”.
• It is a corporate culture that cherishes continuous
improvement.
• SGKM is an implementation strategy for OL.
D. Riaño
Knowledge Management
85
SGKM: Complexity Theory

• Complex adaptive systems (CAS) theory: individuals in a
colony self-organize and continuously fit themselves,
individually and collectively, to changing conditions in
their environment.
D. Riaño
Knowledge Management
86
SGKM: Sustainable Innovation
D. Riaño
Knowledge Management

87
Summary of Terms in KM
•
Balanced Scorecard System (BCS): method of measuring performance of a firm beyond the
typical financial measures. Links corporate goals and direct performance measures in a
framework specific to a firm, and is one method of measuring the impact of knowledge
management. (2)
•
Best Practice: those practices that have produced outstanding results in another situation and
that could be adapted for our situation. (2)
•
Calculated Intangible Value: an "elegant way to put a dollar value on intangible assets" uses a
measure of the company's ability to outperform an average competitor that has similar tangible
assets as the firm's value of intangible assets. Uses the following steps: 1. Calculate average
pretax earnings for three years; 2. Go to the balance sheet and get the average year-end tangible
assets for three years; 3. Divide earnings by assets to get the return on assets. 4. For the same
three years, find the industry's average ROA; 5. Calculate the "excess return" by multiplying the
firm's assets by the industry ROA and subtracting them from the firm's pretax earnings; 6)
calculate the three year average income tax rate and multiply it by the excess return, this results in
the premium attributable to intangible assets; 7) calculate the net present value of the premium by
dividing the premium by the company's cost of capital. (7)
D. Riaño
Knowledge Management
88
Summary of Terms in KM
•
Collaborative Tools: tools such as groupware that enable both structured and free-flow sharing
of knowledge and best practices. An example is Lotus Notes. (2)
•
Communities of Practice: aka affinity groups; A) informal networks and forums, where tips are
exchanged and ideas generated. (7) B) a group of professionals, informally bound to one another
through exposure to a common class of problems, common pursuit of solutions, and thereby
themselves embodying a store of knowledge. (8)
•
Core Capabilities: A) constitute a competitive advantage for a firm; they have built up over time
and cannot be easily imitated. They are distinct from both supplemental and enabling capabilities,
neither of which is sufficiently superior to those of competitors to offer a sustainable advantage.
(6); B) bestow a competitive advantage on a company . . . distinctive, firm-specific, or
organizational competencies; resource deployments; or invisible assets. (6)
•
Core Rigidities: refers to the idea that a firm’s strengths are also – simultaneously – its
weaknesses. The dimensions that distinguish a company competitively have grown up over time
as an accumulation of activities and decisions that focus on one kind of knowledge at the expense
of others. Companies, like people, cannot be skilful at everything. Therefore, core capabilities both
advantage and disadvantage a company. (6)
•
Customer Capital: the value of an organization's relationships with the people with whom it does
business, or the value of its [the companies] franchise, its ongoing relationships with the people or
organizations to which it sells. (7)
D. Riaño
Knowledge Management
89
Summary of Terms in KM
•
Enabling Capabilities: necessary [to enter a market] but not sufficient in themselves to
competitively distinguish a company.
•
Enablers of Knowledge Management: systems and infrastructures which ensure knowledge is
created, captured, shared, and leveraged. These include culture, technology, infrastructure, and
measurement.
•
Experience: refers to what we have done and what has happened to us in the past.
•
Explicit Knowledge: formal/codified . . . comes in the form of books, documents, white papers,
databases, and policy manuals.
•
Human Capital: the capabilities of the individuals required to provide solutions to customers.
•
Intellectual Capital: refers to the commercial value of trademarks, licenses, brand names,
formulations, and patents.
•
Knowledge Interrogators: aka corporate librarian and knowledge integrator; person responsible
for managing the content of organizational knowledge as well as its technology. [they] keep the
database orderly, categorize and format documents and chucking the obsolete, and connect the
users with the information they seek.
D. Riaño
Knowledge Management
90
Summary of Terms in KM
•
Knowledge Management: A) make an organization’s knowledge stores more accessible and
useful; B) a business activity with two primary aspects: treating the knowledge component of
business activities as an explicit concern of business reflected in strategy, policy, and practice at
all levels of the organization making a direct connection between an organization’s intellectual
assets — both explicit [recorded] and tacit [personal know-how] — and positive business results;
C) conscious strategy of getting the right knowledge to the right people at the right time and
helping people share and put information into action in ways that strive to improve organizational
performance.
•
Knowledge Map: representation of the knowledge that exists inside a company.
•
Learning Organization or Organizational Learning: term popularized by Peter Senge's the Fifth
Discipline meaning a corporate culture that cherishes continuous improvement.
•
Market-to-Book Ratio: common method of valuing knowledge intensive companies. Equal to
(price per share X total number of shares outstanding) divided by book equity, which is the equity
portion of a company's balance sheet.
•
Rules of Thumb: shortcuts to solutions to new problems that resemble problems previously
solved by experienced workers.
D. Riaño
Knowledge Management
91
Summary of Terms in KM
•
Signature Skill: an ability by which a person prefers to identify himself or herself professionally.
•
Structural Capital: A) legal rights to ownership: technologies, inventions, data, publications, and
processes [that] can be patented, copyrighted, or shielded by trade-secret laws. B) strategy and
culture, structures and systems, organizational routines and procedures - assets that are often far
more extensive and valuable than the codified ones.
•
Supplemental Capabilities: those that add value to core capabilities but that could be imitated.
•
Technological Capability: term used to encompass the system of activities, physical systems,
skills and knowledge bases, managerial systems of education and reward, and values that create
a special advantage for a company or line of business.
•
Value Proposition: the logical link between action and payoff that knowledge management must
create to be effective. Customer intimacy, product-to-market excellence, and operational
excellence are examples.
D. Riaño
Knowledge Management
92
Summary of Relevant Concepts in KM
Organizational Knowledge (OK): knowledge that is shared among
organizational members. This includes, knowing which information
is needed (know what), knowing how information must be processed
(know how), knowing what information is needed (know why),
knowing where information can be found (know where), and
knowing when which information is needed (know when).
Organizational Learning (OL): “the ability to learn faster than your
competitors”.
Knowledge Map: representation of the knowledge that exists inside a
company.
Knowledge Life Cycle (KLC): knowledge processes as acquisition,
representation, or validation that interact in order to produce new
knowledge.
Knowledge Auditing: identifying which knowledge is needed and
available for that area, which knowledge is missing, who has the
knowledge (source), and how it is being used (destination)”.
Knowledge Process: SGKM term indicating any of the processes
involved in the KM life cycle.
D. Riaño
Knowledge Management
93
KNOWLEDGE STRUCTURES
D. Riaño
Knowledge Management
94
KM Modelling
• Ontologies
•
•
•
•
•
CommonKADS
Protégé 2000
UML-OCL
Multi-Perspective Modelling
Others
D. Riaño
Knowledge Management
95
Ontologies
• Concepts or classes
– Properties or slots
– Facets
Ontology
Constraints
SuperClass
• Relationships
Class
– Inheritance
Class
Property
Facet
• Hierarchy / Network
• Constraints
• Instances
SubClass
Property
Instance
D. Riaño
Knowledge Management
96

CommonKADS
http://www.commonkads.uva.nl
• CommonKADS is a methodology to support structured
knowledge engineering. It is a European de facto
standard for knowledge analysis and knowledge-intensive
system development.
• CommonKADS gives tools for corporate knowledge
management, provides the methods to perform a detailed
analysis of knowledge-intensive tasks & processes, and
supports the development of knowledge systems that
support selected parts of the business process.
• CommonKADS uses UML notations: use case diagrams,
class diagrams, activity diagrams and state diagrams.
D. Riaño
Knowledge Management
97

Protégé 2000
Stanford Medical Informatics, http://protege.stanford.edu
• Java-based standalone application.
• Knowledge Model:
–
–
–
–
–
–
–
–
D. Riaño
Classes: concrete or abstract, hierarchy, multiple-inheritance.
Slots: template or own, hierarchy.
Facets:
Instances:
Meta-classes: classes whose instances are classes.
Forms: screen layouts to edit instances of a class.
Queries: interface for querying the knowledge-base.
PAL constraints:
Knowledge Management
98
Other ontology development tools
•
•
•
•
•
•
•
•
•
•
•
•

APECKS
Apollo
CODE4
Co4
DUET
GKB-Editor
KAO
OilEd
OntoEdit
Visual Ontology Modeler
Unicorn
WebODE
D. Riaño
Knowledge Management
99
Web-based Standards for “Knowledge” Representation
W3C
D. Riaño
Knowledge Management
100
XML: eXtensible Markup Language
•
•
•
•
XML specifies the structure and content of a document.
Extensible: to create a wide variety of document types.
Markup: to increase the description power.
XML is to structure, store and to send information.
Connectivity
Connect
the Web
D. Riaño
Presentation
Browse
the Web
Knowledge Management
Connecting
Applications
Program
the Web
101
HTML …
• HTML was designed for formatting text on a Web page.
• HTML limitations:
–
–
–
–
Cannot deal with the content of a Web page.
Cannot be used to describe or to catalog data in the web.
It is not extensible.
“Standard” representation but browser-dependent appearance.
• HTML browsers supporting XML:
– Microsoft Internet Explorer 5.0
– Netscape Navigator 6 (option “View Page Source”)
• XML reserved symbols: &, <, >, ’, ”, ;.
D. Riaño
Knowledge Management
102
DTD-XML car’s example
<!ELEMENT car
(model?,horsepower,production+,
color?,price?,wheels)
>
<!ATTLIST car
name (BEETLE,SEDAN,JEEP,TOPOLINO,…) #REQUIRED
company IDREF #REQUIRED
>
<!ELEMENT model (#PCDATA)>
<!ELEMENT horsepower (HP|(HPmin?,HPmax?))>
<!ELEMENT production (start?,finish?)>
<!ELEMENT start_prod (#PCDATA)>
<!ELEMENT finish_prod (#PCDATA)>
<!ELEMENT price (factory,retail)>
<!ELEMENT color EMPTY>
<!ATTLIST color name (R|W|B|Y|DARK|OTHER)
#REQUIRED>
<!ELEMENT factory_price (#PCDATA)>
<!ELEMENT retail_price (#PCDATA)>
<!ELEMENT wheels #PCDATA>
<!ELEMENT company (country?)>
<!ATTLIST company ID CDATA>
D. Riaño
<?xml version="1.0"?>
<!DOCTYPE car SYSTEM “car.dtd">
<company ID=“WolksWagen”>
<country>Germany</country>
</company>
…
<car name=“BEETLE” company=“WolksWagen”>
<model>1500</model>
<horsepower>
<HPmin>50</HPmin>
<HPmax>90</HPmax>
</horsepower>
<production>
<start>1938</start>
<finish>1989</finish>
</production>
<production>
<start>2000</start>
<finish>nowadays</finish>
</production>
<color name=“B”></color>
<price>
<factory_price>8000€</fectory_p
rice>
<price>
<wheels>4</wheels>
Knowledge Management
103
</car>
http://www.w3schools.com/default.asp
• eXtensible Stylesheet Language (XSL): transforms XML
into HTML before it is displayed by the browser.
• Document Type Definition (DTD): XML document that
defines the content structure of other XML documents.
• XML Path Language (XPath): locates information in XML
documents.
– Ex. xmlDoc.selectNodes(“//company") selects all the company elements.
– Ex. xmlDoc.selectNodes(“//company[0]") selects the first company element.
– Ex. xmlDoc.selectNodes(“/car") selects all the elements the first company
element.
– Ex. xmlDoc.selectNodes(“/car[color=‘R’]") selects the red car elements.
– Ex. xmlDoc.selectNodes(“/car[@name=‘Beetle’]/horsepower/HPmin”) selects the
HPmin element of all the car with attribute name Beetle.
D. Riaño
Knowledge Management
104
RDF: Resource Description Framework
•
•
•
•
RDF was designed for describing resources on the web.
RDF is to be read and understood by computers
RDF is not for being displayed to people
RDF is written in XML
•
•
RDF uses Uniform Resource Identifiers (URIs).
RDF basic concepts: P(R)=V
– Resources: anything that can have a URI.
– Properties: Resource that has a name.
– Property values: the value of a Property.
•
RDF statements: P(S)=O
– Subject (S): the resource of the statement.
– Predicate (P): the property of the statement.
– Object (O): the property value of the statement.
•
Example: “The webmaster of http://invented.page is John Smith”
Webmaster(http://invented.page)=John Smith
D. Riaño
Knowledge Management
105
RDF Elements
• <rdf:RDF> is the root element of an RDF document.
• <rdf:Description> is the statement constructor that identifies a
resource with the about attribute and contains elements that
describe the resource.
• <rdf:Bag> describes a list of values that is intended to be unordered.
Ex. <rdf:Description rdf:about="http://www.old_cars.org">
<car>
<rdf:Bag>
Car (www.old_cars.org)=
<rdf:li>Beetle</rdf:li>
(Beetle, Sedan, Jeep, …)
<rdf:li>Sedan</rdf:li>
<rdf:li>Jeep</rdf:li>
…
<rdf:Bag>
</car>
</rdf:Description>
• <rdf:Seq> describes a list of values that is intended to be ordered.
• <rdf:Alt> describes a list of alternative values.
D. Riaño
Knowledge Management
106
RDF Schema (RDFS)
•
•
RDFS provides the framework to describe application-specific classes and properties.
Classes in RDFS allows resources to be defined as instances of classes.
Ex.
<?xml version="1.0"?>
<rdf:RDF
xmlns:rdf= "http://www.w3.org/1999/02/22-rdf-syntax-ns#"
xmlns:rdfs="http://www.w3.org/2000/01/rdf-schema#"
xml:base= "http://www.animals.fake/cars#">
SubclassOf
CAR
OLD_CAR
<rdf:Description rdf:ID="car">
<rdf:type
rdf:resource="http://www.w3.org/2000/01/rdf-schema#Class"/>
</rdf:Description>
<rdf:Description rdf:ID="old_car">
<rdf:type
rdf:resource="http://www.w3.org/2000/01/rdf-schema#Class"/>
<rdfs:subClassOf rdf:resource="#car"/>
</rdf:Description>
</rdf:RDF>
D. Riaño
Knowledge Management
107

DAML
D. Riaño
Knowledge Management
108

OIL
D. Riaño
Knowledge Management
109

DAML-OIL
D. Riaño
Knowledge Management
110
OWL: Web Ontology Language
•
•
•
•
OWL is built on top of RDF and written in XML.
OWL is for processing information on the web
OWL was designed to be interpreted by computers and not for being read by people
OWL has three sublanguages
–
–
–
•
OWL DL  OWL full within DL fragment
DL semantics officially definitive
OWL DL based on SHIQ Description Logic
–
•
OWL full OWL syntax + RDF (complete expressiveness without computational guarantees)
OWL DL restricted to FOL fragment (computational complete & decidable reasoning K)
OWL Lite is “easier to implement” subset of OWL DL (hierarchical K)
Semantic layering
–
–
•

In fact it is equivalent to SHOIN(Dn) DL
OWL DL Benefits from many years of DL research
–
–
–
–
D. Riaño
Well defined semantics
Formal properties well understood (complexity, decidability)
Known reasoning algorithms
Implemented systems (highly optimised)
Knowledge Management
111
OWL: Class Constructors and Axioms
D. Riaño
Knowledge Management
112
OWL: Example
Person hasChild.Doctor
 hasChild.Doctor
<owl:Class>
<owl:intersectionOf rdf:parseType=" collection">
<owl:Class rdf:about="#Person"/>
<owl:Restriction>
<owl:onProperty rdf:resource="#hasChild"/>
<owl:toClass>
<owl:unionOf rdf:parseType=" collection">
<owl:Class rdf:about="#Doctor"/>
<owl:Restriction>
<owl:onProperty rdf:resource="#hasChild"/>
<owl:hasClass rdf:resource="#Doctor"/>
</owl:Restriction>
</owl:unionOf>
</owl:toClass>
</owl:Restriction>
</owl:intersectionOf>
</owl:Class>
D. Riaño
Knowledge Management
113

KM-CMM
•
•
•
•
•
•
Knowledge Management Capability Maturity Model.
CMM: According to the Carnegie Mellon University Software Engineering Institute, CMM is a
common-sense application of software or business process management and quality
improvement concepts to software development and maintenance. Its a community-developed
guide for evolving towards a culture of engineering excellence, model for organizational
improvement. The underlying structure for reliable and consistent software process assessments
and software capability evaluations.
The Capability Maturity Model for Software (CMM) is a framework that describes the key elements
of an effective software process. There are CMMs for non software processes as well, such as
Business Process Management (BPM). The CMM describes an evolutionary improvement path
from an ad hoc, immature process to a mature, disciplined process. The CMM covers practices for
planning, engineering, and managing software development and maintenance. When followed,
these key practices improve the ability of organizations to meet goals for cost, schedule,
functionality, and product quality. The CMM establishes a yardstick against which it is possible to
judge, in a repeatable way, the maturity of an organization's software process and compare it to
the state of the practice of the industry. The CMM can also be used by an organization to plan
improvements to its software process. It also reflects the needs of individuals performing software
process, improvement, software process assessments, or software capability evaluations; is
documented; and is publicly available.
Intro CMM: http://www.dis.wa.gov/portfolio/tr25/tr25_o2.html
P-CMM: http://www.sei.cmu.edu/cmm-p/
CMMI:
D. Riaño
Knowledge Management
114
KNOWLEDGE TECHNOLOGIES
D. Riaño
Knowledge Management
115

KM technologies & tools
TECHNOLOGIES
• Management Sciences
• Artificial Intelligence
• Information Retrieval
• Organizational Behaviour
D. Riaño
Google Directory TOOLS
• Brainstorming (10)
• Business Intelligence (13)
• Classification (18)
• Collaboration (31)
• Concept Mapping (6)
• Data Mining (89)
• Information Retrieval (119)
• Knowledge Discovery (34)
• Online Training Systems (93)
• Topic Maps (93)
Knowledge Management
116

KM Technologies
• Management Sciences
• Artificial Intelligence
–
–
–
–
–
–
–
Case-Based Reasoning
Ontology-Based KM
Metadata-Based KM
Knowledge Discovery
Knowledge Acquisition
Data-, Text-, and Web- Mining
Intelligent Agents
• Information Retrieval
– Information Retrieval&Extraction
– Visualisation Techniques:
• Organizational Behaviour
D. Riaño
Knowledge Management
117
Case-Based Reasoning
D. Riaño
Knowledge Management
118

Data Mining
•
•
•
Corpus
Pre-processing + Data Mining + Analysis
Sort of data:
–
–
–
•
Data mining technologies
–
–
–
–
–
•
•
Structured data: databases, XML, RDF, DAML-OIL, OWL, etc.
Semi-structured data: HTML-like documents.
Non-structured data: Textual documents.
Artificial Neural Networks: modelos predecible no-lineales que aprenden a través del
entrenamiento y semejan la estructura de una red neuronal biológica.
Decision Trees: estructuras de forma de árbol que representan conjuntos de decisiones.
Estas decisiones generan reglas para la clasificación de un conjunto de datos. Métodos
específicos de árboles de decisión incluyen Arboles de Clasificación y Regresión (CART:
Classification And Regression Tree) y Detección de Interacción Automática de Chi Cuadrado
(CHAI: Chi Square Automatic Interaction Detection)
Genetic Algorithms: técnicas de optimización que usan procesos tales como
combinaciones genéticas, mutaciones y selección natural en un diseño basado en los
conceptos de evolución.
Nearest Neighbour: una técnica que clasifica cada registro en un conjunto de datos basado
en una combinación de las clases del/de los k registro (s) más similar/es a él en un conjunto
de datos históricos (donde k
1). Algunas veces se llama la técnica del vecino k-más
cercano.
Inductive Rules: la extracción de reglas if-then de datos basados en significado estadístico.
Text mining.
Web mining.
D. Riaño
Knowledge Management
119
Text Mining
•
•
•
•
•
•
Authority file: list of important words in a domain or area of expertise.
Equation (1): Inverse document frequency (Spark Jones, 1970).
Equation (2): Weight of a term that appears in n out of N documents.
Equation (3): Relative weight of a term being p the probability that the terms
appears in a relevant document, and q the probability that it appear in an
irrelevant document.
Equation (4): Relative weight of a term that appears in r out of R relevant
documents, and in n out of N non relevant documents. =0,5 is defined to
avoid divide-by-zero problems.

WordNet
N
n
N
n
p(1  q)
  log
q(1  p)
(r   )( N  R  n  r   )
  log
( R  r   )( n  r   )
  log
D. Riaño
Knowledge Management
(1)
(2)
(3)
(4)
120

Web Mining: metrics (1)
(Dhyani, Ng & Bhowmick, 2002)
• Graph properties
– Centrality
– Global
– Local
ODi   C ij
IDi   C ji
j
ROC i 
 C
i
ij
RIC i 
i
ji
j
C
ji
j
• Significance
– Relevance
– Quality


I ij  c 2


Max  Min
i
j
ji
j
LAP
 N3
if n is even

LAP   3 4
 N  N otherwise
 4
c1
if X ij  1
if  k : 0  c  1 : X kj  1 & Li kj  Lo kj  0
0 otherwise
M
Riq 
I
j 1
Riq 
ij
( Boolean spread activation)
M
M


 Liik  X kj  ( Most cited )


k 1, k  i 
j 1

TFij
(0.5  0.5 
)  IDF j

max k 1.. M {TFik }
Qj
Riq 
N

 (0.5  0.5  max
jPi
D. Riaño
[0]
if a has no child

D(a)  
otherwise
[1  Max ( D(a1 )), ,1  Max ( D(a n ))]
[
0
]
if
a
has
no child

C (a)  
otherwise
{1   C (a1 ), ,1   C (a n )}
j
ij
S
j
i
 C  C
ij
j
C
 C
Max   C ij
Cp 
j
TFij
k 1.. M
{TFik }
)  IDF
2
Knowledge Management
(TFxIDF )
2
j
121
Web Mining: metrics (2)
• Similarity
r ( a, b) 
– Content
– Link
c ( a, b) 
Sa  Sb
Sa  Sb
S a  Sb
S ij   TFik  TF jk
• Search
Sa
(resemblance)
(containment )
(Term  based similarity )
k
CiT  (CiT ) T
(co  citation strength)
C  U T  C Tj  U T
1
S ijs  lij
(direct path strength)
l
2  2 ij
1
S ija   j
(common ancestor strength)
lkji
l
kAij 2 ki  2
1
S ijd   j
(common descendents strength)
l ijk
l
kDij 2 ik  2
S ijC 
T
i
– Effectiveness
• Precision
• Recall
– Comparison
• Usage
• Information theoretic
D. Riaño
Knowledge Management
122
Intelligent Agents
D. Riaño
Knowledge Management
123
KNOWLEDGE TOOLS
D. Riaño
Knowledge Management
124

KM Tools
•
•
•
•
•
•
•
Knowledge
Knowledge
Knowledge
Knowledge
Knowledge
Knowledge
Others:
–
–
–
–
–
–
–
–
–
–
D. Riaño
capture: Clementine
access: AQUAINT
mining:
summarization:
mapping:
visualization:
Brainstorming
Business Intelligence
Classification
Collaboration
Concept Mapping
Data Mining
Information Retrieval
Knowledge Discovery
Online Training Systems
Topic Maps
Knowledge Management
125
Clementine
•
•
•
•
•
•
•
•
•
•
•
•
www.spss.com/clementine
Visual Programming Interface
builds a discovery model
performs learning task
Uses neural networks and rule induction
Data sources
ASCII file format, Oracle, Informix, Sybase and Ingres
Clementine has many useful facilities:
Data Manipulation - construct new data items derived from existing ones,
and breaking the data down into meaningful sub-sets
Browsing and Visualisation - displaying aspects of the data using interactive
graphics
Statistics - confirming suspected relationships between factors in the data
Hypothesis testing - constructing models of how the data behaves and
verifying them
D. Riaño
Knowledge Management
126

AQUAINT
• Advanced Question Answering for Intelligence
• www.ic-arda.org/InfoExploit/aquaint
D. Riaño
Knowledge Management
127
Excalibur
• Excalibur RetrievalWare delivers advanced knowledge
retrieval solutions for the full spectrum of digital
information. Excalibur's semantic networks and Adaptive
Pattern Recognition Processing provide highly faulttolerant fuzzy searching and plain English meaningbased searching for text, and powerful query-by-example
searching for multimedia.
D. Riaño
Knowledge Management
128
Google
D. Riaño
Knowledge Management
129
EuroSpider
• The EUROSPIDER system is an Information Retrieval (IR) system
which searches very large and complex data collections for relevant
information. It is a commercial version of the IR system SPIDER,
developed by the Swiss Federal Institute of Technology.
EUROSPIDER
can
be
used
in
various
ways:
1.
as
a
standalone
IR
system
2. as an add-on to a World-Wide Web server which makes data
collection accessible through a private or public network
3. added to a commercial database (DB) system to access possibly
very
dynamic
and
structured
data.
The EUROSPIDER retrieval system provides advanced Information
Retrieval (IR) functions such as relevance ranking, feedback
searches, linguistic document analysis, and automatic indexing.
Document analysis and indexing optionally includes fuzzy term
matching to cope with recognition errors of OCR-devices.
D. Riaño
Knowledge Management
130
hTechSight
D. Riaño
Knowledge Management
131
GATE
• GATE is open source Java software under the GNU library licence,
and is a stable, robust, and scalable infrastructure which allows
users to build and customise language processing components,
while mundane tasks like data storage, format analysis and data
visualisation are handled by GATE. The system is bundled with
components for language analysis, and is in use for Information
Extraction (IE), Information Retrieval (IR), Natural Language
Generation, summarisation, dialogue, Semantic Web, Knowledge
Technologies and Digital Libraries applications. GATE-based
systems have taken part in the all the major quantitative evaluation
programmes for Natural Language Processing since 1995.
D. Riaño
Knowledge Management
132
Knowledge Validation, Verification, and Testing

• Validation
• Verification
• Testing automated systems: wizard-oz experiment,
simulation.
D. Riaño
Knowledge Management
133

DM tools
• SPSS - Clementine
– http://www.spss.com/clementine/
• Oracle - Darwin
– http://www.oracle.com/ip/analyze/warehouse/datamining/
• SGI - MineSet
– http://www.sgi.com/software/mineset/
• IBM - Intelligent Miner
– http://www-4.ibm.com/software/data/iminer/fordata/
• http://www.kdnuggets.com/software/index.html
D. Riaño
Knowledge Management
134
KNOWLEDGE ENGINERING
D. Riaño
Knowledge Management
135
Knowledge Engineering Life Cycle
Purpose: generate K bases and knowledge-based systems
1.
2.
3.
4.
5.
6.
7.
Problem selection
Knowledge acquisition
Knowledge representation
Knowledge encoding
Knowledge testing and evaluation
If more refinement is required, then go to 2
Implementation and maintenance
D. Riaño
Knowledge Management
136
The SWOT Analysis
•
•
•
Methodology for scanning the internal and external environment of a
firm or process.
Used as an important part of the strategic planning process and also
for analyzing KM systems.
Strengths, Weaknesses, Opportunities, and Threads analysis.
– Strengths (S): company resources and capabilities that can be used as a basis
for developing a competitive advantage. For example, patents, strong brand
names, good reputation among customers, cost advantages from proprietary
know-how, exclusive access to high grade natural resources, favorable access to
distribution networks.
– Weaknesses (W): The absence of certain strengths may be viewed as a
weakness. For example, lack of patent protection, a weak brand name, poor
reputation among customers, high cost structure , lack of access to the best
natural resources , lack of access to key distribution channels.
– Opportunities (O): new opportunities for company profit and growth. For
example, an unfulfilled customer need, arrival of new technologies, loosening of
regulations, removal of international trade barriers.
– Threads (T): changes in the external environmental that may represent threats
to the firm. For example, shifts in consumer tastes away from the firm's products,
emergence of substitute products, new regulations, increased trade barriers.
D. Riaño
Knowledge Management
137
The SWOT matrix: strategies
MATRIX
Opportunities
Threats
D. Riaño
Strengths
Weaknesses
S-O strategies
W-O strategies
pursue opportunities that overcome weaknesses to
are a good fit to the pursue opportunities.
companies strengths.
S-T strategies
identify ways that the firm
can use its strengths to
reduce its vulnerability to
external threats.
W-T strategies
establish a defensive plan
to
prevent
the
firm's
weaknesses from making it
highly
susceptible
to
external threats.
Knowledge Management
138
GOAL-ORIENTED METHODOLOGIES
D. Riaño
Knowledge Management
139
Goal-Oriented
• Decision Making
• Semantic Web
• Agents
D. Riaño
Knowledge Management
140
Decision Making in the context of KM
•
•
•
•
•
•
Decision Trees
Decision Graphs
Decision Tables
Rules
Influence Diagrams
Bayesian Networks
D. Riaño
Knowledge Management
141

Decision Trees
• (Hunt, Marin & Stone, 1966) A decision tree is a tree structure
consisting of decision nodes and leaves. Decision nodes specify an
attribute to test upon an object, with the arcs out of the decision
node specifying the possible values that attribute can take. Each leaf
of the decision tree specifies a category in the set of possible
decisions.
• Example:
• Properties:
– Intuitive and easy to use, implement, automate, etc.
– Production rules equivalent
– The replication and the fragmentation problems.
• (production) Decision tree induction: ID3, C4.5, etc.
D. Riaño
Knowledge Management
142

Decision Graphs
• (Oliver, 1993) A decision graph is a generalization of a decision tree
having decision nodes, decision leaves, and joins. A join is
represented as a set of nodes having a common child.
• Example:
• Properties:
– Do not have the replication and fragmentation problems.
– Difficult to make it equivalent to decision rules.
– Difficult to automate.
D. Riaño
Knowledge Management
143

Decision Tables
•
•
•
•
Definición
Extended decision tables
Example:
Guías de práctica clínica
D. Riaño
Knowledge Management
144
Rules
• IF condition THEN conclusion
• Example: if a project on market analysis is required, then make the
project manager to have a marketing profile.
• Sorts of rules:
– Production rules: IF condition THEN concept
(ex. IF sales > 1$ million THEN 1st_class_seller)
– Association rules: IF (x1,…,xk)=(v1,…,vk) THEN (y1,…,yj)=(w1,…,wj)
(ex. IF (sort,seniority)=(1st_class_seller,15 years) THEN salary_incr=15%)
– Ripple down rules (RDR): IF condition THEN conclusion EXCEPT RDR
(ex. IF seniority=15 years THEN salary_incr=10% EXCEPT IF
sort=1st_class_seller THEN salary_incr=15% ELSE salary_incr=5%)
• (production) Rule Induction: AQ algorithms, CN2, etc.
• (use) Inference Engine: forward & backward chainning.
D. Riaño
Knowledge Management
145

Influence Diagrams
• Definition
• Node types:
– Utility
– Decision
– ?
D. Riaño
Knowledge Management
146

Bayesian Networks
• Definition
• Bayesian Probability Theory
• Example
D. Riaño
Knowledge Management
147
Semantic Web in the context of KM
D. Riaño
Knowledge Management

148

From Syntactic Web to Semantic Web
• What is new with semantics?
– Complex queries involving background knowledge
• Find information about “animals that use sonar but are not
either bats or dolphins”
– Locating information in data repositories
• Travel enquiries
• Prices of goods and services
• Results of human genome experiments
– Finding and using “web services”
• Visualise surface interactions between two proteins
– Delegating complex tasks to web “agents”
• Book me a holiday next weekend somewhere warm, not too
far away, and where they speak French or English
D. Riaño
Knowledge Management
149
Agents in the context of KM
Information
Retrieval
Distributed
Systems
Mobile code
AI & Cognitive
Science
agents
Machine
Learning
Database &
Knowledge base
Technology
agents
2003
D. Riaño
=
objects
1982
structured
programming
= 1974
Knowledge Management
150

Agent Architecture
• BDI
Knowledge
Reflection
BDI
Knowledge
Base
communication
D. Riaño
Knowledge Management
151
Multi-Agent Systems
D. Riaño
Knowledge Management

152
Multi-Agent System Engineering
(DeLoach, 1999)
1.
2.
3.
4.
Identify the sort of agents
Identify the interaction between these agents
Define the coordination protocols for each interaction
Map the actions fired during the conversations into
agent internal components.
5. Define the input, flows, and outputs of the agents
6. Select the sorts of agents required.
7. Determine the physical location of the agents and other
possible parameters of the agents.
D. Riaño
Knowledge Management
153
Multi-Agent Platforms

• JADE
D. Riaño
Knowledge Management
154
Multi-Agent Knowledge Networks: an example.
H-TechSight
D. Riaño
Knowledge Management
155
References
•
•
•
•
•
•
Davenport TH, Prusak L. Working Knowledge. Harvard Business School Press, 2000.
Liebowitz J. Knowledge Management. CRC Press, 2001.
Liebowitz J. Knowledge Management Handbook. CRC Press 1999.
McElroy MW. The New knowledge management. Butterworth-Heinemann, 2003.
Bañares-Alcantara R. KM and AI course. University of Oxford, 2004.
Dhyani D., Ng W.K., Bhowmick S.V. A Survey of Web Metrics. ACM Comp. Surveys
24(4), 2002.
D. Riaño
Knowledge Management
156