Topic Maps

Steiner, Knud

Topic Maps

An Enabling Technology for Knowledge Management and its Implementation

von Knud Steiner (Autor:in)

BWL - Unternehmensführung, Management, Organisation

Zusammenfassung

Inhaltsangabe:Einleitung:
Der Status von Wissen als wertvolle und strategisch bedeutsame Ressource von Organisationen ist unbestritten. Von einem technologischen Standpunkt aus betrachtet, ist der Aufbau einer organisationalen Wissensbasis, also die Identifizierung, Erschließung, Entwicklung, Verbreitung, (Wieder-)Verwendung und Bewahrung von organisationalem Wissen, bestmöglich durch den Einsatz von Informations- und Kommunikationstechnologie zu unterstützen. Im Idealfall geschieht das durch Anwendungssysteme (sog. "Wissensmanagementsysteme"), mit deren Hilfe die Erfassung von sowohl explizitem als auch implizitem Wissen, seine Speicherung in digitalen Wissensobjekten und -einheiten, deren systematische Ordnung (z.B. durch Kategorisierung), und deren kontextbezogene Verknüpfung und Bereitstellung bewerkstelligt werden kann.
Zwei etablierte Ansätze für Wissensmanagementsysteme sind der auf sog. "Superimposed Information" basierende Ansatz und der ontologiebasierte Ansatz. Superimposed Information überlagert bereits vorhandene explizite Wissensressourcen (z.B. elektronische Dokumente) mit dem Zweck, deren Erschließung qualitativ zu verbessern, ohne dabei die Struktur und den Inhalt dieser Ressourcen zu verändern (z.B. durch Klassifizierung und/oder Kommentierung mit Metadaten). Ontologien werden hingegen häufig dazu verwendet, noch nicht explizit vorhandenes organisationales Wissen zu formalisieren (z.B. in Form von Taxonomien).
In vorliegender Arbeit wird das mit ISO/IEC 13250 standardisierte abstrakte Modell fokussiert und aufgezeigt, dass Topic Maps - semantische Netze als Instanzen dieses Modells - sowohl für die Repräsentation von Superimposed Information als auch für die Repräsentation von Ontologien geeignet sind. Ferner wird demonstriert, dass Topic Maps eingesetzt werden können, um beide Ansätze für den evolutionären Aufbau von elektronischen organisationalen Wissensbasen zu kombinieren.
Als anwendungsorientiertes Ergebnis vorliegender Arbeit wird ein Ansatz zur persistenten Speicherung und Wartung von Topic Maps vorgestellt, der auf relationaler Datenbanktechnologie basiert. Ein wesentliches Merkmal dieser "Topic Map Engine" ist die Implementierung von Mechanismen, mit deren Hilfe ein sog. "Topic Map Schema" (ein "Topic Map Template" und eine Menge von grundlegenden semantischen Konsistenz- und Gültigkeitsbedingungen) verwaltet werden kann. Obwohl solche Mechanismen sowohl für den Entwurf als auch für die Wartung von umfangreichen und […]

Leseprobe

Inhaltsverzeichnis

ID 5022

Steiner, Knud: Topic Maps: An Enabling Technology for Knowledge Management and its

Implementation / Knud Steiner - Hamburg: Diplomica GmbH, 2002

Zugl.: Linz, Universität, Diplom, 2001

Dieses Werk ist urheberrechtlich geschützt. Die dadurch begründeten Rechte, insbesondere die

der Übersetzung, des Nachdrucks, des Vortrags, der Entnahme von Abbildungen und Tabellen,

der Funksendung, der Mikroverfilmung oder der Vervielfältigung auf anderen Wegen und der

Speicherung in Datenverarbeitungsanlagen, bleiben, auch bei nur auszugsweiser Verwertung,

vorbehalten. Eine Vervielfältigung dieses Werkes oder von Teilen dieses Werkes ist auch im

Einzelfall nur in den Grenzen der gesetzlichen Bestimmungen des Urheberrechtsgesetzes der

Bundesrepublik Deutschland in der jeweils geltenden Fassung zulässig. Sie ist grundsätzlich

vergütungspflichtig. Zuwiderhandlungen unterliegen den Strafbestimmungen des

Urheberrechtes.

Die Wiedergabe von Gebrauchsnamen, Handelsnamen, Warenbezeichnungen usw. in diesem

Werk berechtigt auch ohne besondere Kennzeichnung nicht zu der Annahme, dass solche

Namen im Sinne der Warenzeichen- und Markenschutz-Gesetzgebung als frei zu betrachten

wären und daher von jedermann benutzt werden dürften.

Die Informationen in diesem Werk wurden mit Sorgfalt erarbeitet. Dennoch können Fehler nicht

vollständig ausgeschlossen werden, und die Diplomarbeiten Agentur, die Autoren oder

Übersetzer übernehmen keine juristische Verantwortung oder irgendeine Haftung für evtl.

verbliebene fehlerhafte Angaben und deren Folgen.

Diplomica GmbH

http://www.diplom.de, Hamburg 2002

Printed in Germany

Wissensquellen gewinnbringend nutzen

Qualität, Praxisrelevanz und Aktualität zeichnen unsere Studien aus. Wir

bieten Ihnen im Auftrag unserer Autorinnen und Autoren Wirtschafts-

studien und wissenschaftliche Abschlussarbeiten Dissertationen,

Diplomarbeiten, Magisterarbeiten, Staatsexamensarbeiten und Studien-

arbeiten zum Kauf. Sie wurden an deutschen Universitäten, Fachhoch-

schulen, Akademien oder vergleichbaren Institutionen der Europäischen

Union geschrieben. Der Notendurchschnitt liegt bei 1,5.

Wettbewerbsvorteile verschaffen Vergleichen Sie den Preis unserer

Studien mit den Honoraren externer Berater. Um dieses Wissen selbst

zusammenzutragen, müssten Sie viel Zeit und Geld aufbringen.

http://www.diplom.de bietet Ihnen unser vollständiges Lieferprogramm

mit mehreren tausend Studien im Internet. Neben dem Online-Katalog und

der Online-Suchmaschine für Ihre Recherche steht Ihnen auch eine Online-

Bestellfunktion zur Verfügung. Inhaltliche Zusammenfassungen und

Inhaltsverzeichnisse zu jeder Studie sind im Internet einsehbar.

Individueller Service

Gerne senden wir Ihnen auch unseren Papier-

katalog zu. Bitte fordern Sie Ihr individuelles Exemplar bei uns an. Für

Fragen, Anregungen und individuelle Anfragen stehen wir Ihnen gerne zur

Verfügung. Wir freuen uns auf eine gute Zusammenarbeit.

Ihr Team der Diplomarbeiten Agentur

Kurzfassung

Der Status von Wissen als wertvolle und strategisch bedeutsame Ressource von Organisationen

ist unbestritten. Von einem technologischen Standpunkt aus betrachtet, ist der Aufbau einer

organisationalen Wissensbasis, also die Identifizierung, Erschließung, Entwicklung, Verbreitung,

(Wieder-)Verwendung und Bewahrung von organisationalem Wissen, bestmöglich durch den

Einsatz von Informations- und Kommunikationstechnologie zu unterstützen. Im Idealfall

geschieht das durch Anwendungssysteme (sog. ,Wissensmanagementsysteme`), mit deren Hilfe

die Erfassung von sowohl explizitem als auch implizitem Wissen, seine Speicherung in digitalen

Wissensobjekten und -einheiten, deren systematische Ordnung (z.B. durch Kategorisierung), und

deren kontextbezogene Verknüpfung und Bereitstellung bewerkstelligt werden kann.

Zwei etablierte Ansätze für Wissensmanagementsysteme sind der auf sog. ,Superimposed

Information` basierende Ansatz und der ontologiebasierte Ansatz. Superimposed Information

überlagert bereits vorhandene explizite Wissensressourcen (z.B. elektronische Dokumente) mit

dem Zweck, deren Erschließung qualitativ zu verbessern, ohne dabei die Struktur und den Inhalt

dieser Ressourcen zu verändern (z.B. durch Klassifizierung und/oder Kommentierung mit

Metadaten). Ontologien werden hingegen häufig dazu verwendet, noch nicht explizit

vorhandenes organisationales Wissen zu formalisieren (z.B. in Form von Taxonomien).

In vorliegender Arbeit wird das mit ISO/IEC 13250 standardisierte abstrakte Modell fokussiert

und aufgezeigt, dass Topic Maps -- semantische Netze als Instanzen dieses Modells -- sowohl

für die Repräsentation von Superimposed Information als auch für die Repräsentation von

Ontologien geeignet sind. Ferner wird demonstriert, dass Topic Maps eingesetzt werden können,

um beide Ansätze für den evolutionären Aufbau von elektronischen organisationalen

Wissensbasen zu kombinieren.

Als anwendungsorientiertes Ergebnis vorliegender Arbeit wird ein Ansatz zur persistenten

Speicherung und Wartung von Topic Maps vorgestellt, der auf relationaler

Datenbanktechnologie basiert. Ein wesentliches Merkmal dieser ,Topic Map Engine` ist die

Implementierung von Mechanismen, mit deren Hilfe ein sog. ,Topic Map Schema` (ein ,Topic

Map Template` und eine Menge von grundlegenden semantischen Konsistenz- und

Gültigkeitsbedingungen) verwaltet werden kann. Obwohl solche Mechanismen sowohl für den

Entwurf als auch für die Wartung von umfangreichen und komplexen Topic Maps als

unabdingbar angesehen werden können, sind entsprechende Konzepte in der aktuellen Version

von ISO/IEC 13250 nicht spezifiziert.

Abstract

The treatment of organizational knowledge as a valuable strategic asset and its management is

becoming widely accepted. From a technological point of view, the identification, acquisition,

development, dissemination, (re-)use, and preservation of an organization's knowledge are to be

optimally supported with information and communication technologies, mainly by means of or-

ganizational memory (or knowledge management) systems. Ideally, such systems capture

knowledge, define, store, categorize, index, and link digital objects corresponding to knowledge

units, and provide an interface for searching for (`pulling') and subscribing to (`pushing') rele-

vant knowledge that is presented with sufficient flexibility to render it meaningful and applicable

across multiple contexts of use.

In this thesis, the author shows how Topic Maps (semantically structured, self-describing link

networks standardized by ISO/IEC 13250) can be used to represent superimposed information as

well as domain ontologies -- two established approaches to organizational memory systems. Su-

perimposed information enriches explicit knowledge resources (e.g. electronic documents) for

purposes like retrieval or connection, without modifying this base information. Domain ontolo-

gies are frequently used to capture and formalize former implicit domain-specific knowledge.

Since the underlying abstract model of topic maps provides a high degree of power and flexibil-

ity, topic maps can also be used to combine both approaches as a framework supporting the evo-

lutionary construction of computer-based organizational memories that grow both in structure

and extent.

The applicative outcome of this thesis is an approach towards the persistent storage and mainte-

nance of large and complex topic maps with relational database technology. A major feature of

this `Topic Map Engine' is that it provides support for the specification of a so-called `topic map

schema' (i.e. a `topic map template' and a set of basic constraints) using native concepts of the

topic map model. Although topic map schemas are an indispensable requirement for the design

and maintenance of semantically valid topic maps, they are not covered by ISO/IEC 13250 in its

current version.

List of Abbreviations

AI Artificial Intelligence

API Application Programming Interface

CKBS Competence Knowledge Base System

CM Corporate Memory

CSCL Computer Supported Cooperative Learning

CSCW Computer Supported Cooperative Work

DML Data Manipulation Language

DMS Document Management System

DTD Document Type Definition

EIP Enterprise Information Portal

EKP Enterprise Knowledge Portal

ERM Entity Relationship Model

ICT Information and Communication Technology

IM Information Management

KM Knowledge Management

KMS Knowledge Management System

OM Organizational Memory

OMS Organizational Memory System

OODB Object Oriented Database

OOM Object Oriented Model

PL/SQL Procedural Language/SQL

PSI Public Subject Indicator

RDB Relational Database

RDBMS Relational Database Management System

RDF Resource Description Framework

SA Software Agent

SGML Standard Generalized Markup Language

SQL Structured Query Language

TME Topic Map Engine

UML Unified Modeling Language

URI Uniform Resource Identifier

XML Extensible Markup Language

XTM XML Topic Maps

Contents

1 Introduction ...1

1.1 Structure...2

2 Knowledge and Knowledge Management...3

2.1 Knowledge...3

2.1.1 Types of Knowledge ...4

2.2 Knowledge Management...5

2.2.1 An Interdisciplinary Field of Research ...6

2.2.2 Knowledge Work...8

2.2.3 `Knowing' Organizations ...10

2.2.4 Organizational Memory...12

2.2.5 Organizational Point of View...15

2.2.6 Technological Point of View...19

2.2.7 Knowledge Management vs. Information Management...25

3 The Topic Map Model...29

3.1 Introduction and Background...29

3.2 Topic Map and Topic Map Document...31

3.3 Key Concepts...33

3.3.1 Topic and Topic Type ...33

3.3.2 Topic Names...34

3.3.3 Topic Occurrence and Occurrence Role Type...36

3.3.4 Topic Association, Association Type, and Association Role Type ...37

3.4 Advanced Concepts ...38

3.4.1 Scope and Theme ...38

3.4.2 Facet, Facet Type, and Facet Value Type...42

3.4.3 Public Subject ...44

4 Contributions of the Topic Map Model to Organizational Memory

Systems ...45

4.1 Representation of Superimposed Information...45

4.1.1 Topic Map Schema...48

4.1.2 Topic Maps vs. the Resource Description Framework...53

4.2 Representation of Domain Ontologies...56

4.2.1 Basic Association Types...58

4.2.2 Association Properties ...60

4.3 Evolutionary Construction of Organizational Memories with Topic Maps ..61

4.3.1 Early Evolution Stage ...61

4.3.2 Ongoing Evolution Stage ...62

5 Design and Implementation of a Relational DatabaseBased Topic

Map Engine ...64

5.1 Motivation...64

5.2 Overall System Architecture...66

5.3 Data Models...67

5.3.1 Conceptual Data Model ...68

5.3.2 Logical Data Model ...69

5.3.3 Relational Database Schema ...70

5.4 Modules ...81

5.4.1 Core Module ...81

5.4.2 Schema Module...83

5.4.3 Instance Module ...87

5.5 Application Programming Interfaces ...90

5.5.1 Core Interface ...90

5.5.2 Topic Map Designer's Interface ...95

5.5.3 Topic Map Editor's Interface ...101

6 Conclusions and Related Work ...106

Appendix A: Sample Topic Map Instance ...108

Appendix B: Sample Topic Map Schema ...110

List of Figures...112

References ...113

1 Introduction

The treatment of organizational knowledge as a valuable strategic asset and its management is

becoming widely accepted. Business organizations within a broad range of industries are moving

towards improved ways and means of handling their knowledge in general, especially its crea-

tion and preservation within the organization, and the sharing of knowledge among its members.

The main goal of these interventions is to cope with the following changes in the business world

[e.g. Stein, 1995]:

· the increased complexity, dynamics, fragmentation, and decentralization of knowledge or

knowledge development,

· the increased complexity of organizational structures and the permanent need to change

them, and

· the increased amount of non-traditional data to be managed, such as hypertext documents,

links, multimedia documents, and communication acts.

From a technological point of view, the management of organizational knowledge can be basi-

cally supported by application systems that facilitate the codification of knowledge, and that

provide access to it by means of advanced information retrieval techniques going beyond state-

of-the-art full-text search.

Both research prototypes and commercial systems designed for these purposes are mainly based

on the following two approaches:

· the overlay of already existing resources of codified knowledge (mainly electronic docu-

ments in various formats) with `value-adding' meta-data (or `superimposed information') in

order to refine the retrieval process (e.g. location, annotation, classification) and

· the use of semantic networks as building blocks of Artificial Intelligence (AI) applications,

such as inference engines and expert systems, for the formal representation of domain-

specific knowledge capturing a shared and common understanding of a specific knowledge

domain and enabling its retrieval with advanced information and knowledge retrieval tech-

niques.

As a major drawback of these systems, however, the abstract models of their underlying knowl-

edge structures are left to individual implementations, thus making it difficult, for example, to

develop unified models for the conceptual design of such structures, to implement common-

purpose tools for their creation, and to interchange them between different application systems.

Therefore, a need for standardized models and interchange formats that are flexible and powerful

enough to support the conceptual design, representation, and interchange of semantically struc-

tured meta-data and/or semantic networks is obvious.

The abstract model defined by the ISO standard ISO/IEC 13250:2000 Topic Maps [ISO13250,

2000] aims at being a major choice for these purposes in future. This model has already attracted

a lot of interest, since it is very powerful and flexible and allows the representation and inter-

change of knowledge and navigational structures, that has not been possible in a standardized

way up to present.

1.1 Structure

The structure of this thesis is as follows:

· Chapter 2: Knowledge and Knowledge Management

gives a compact introduction to the management of knowledge -- the general topic of inter-

est within the scope of this thesis -- based on a working definition for `Knowledge Man-

agement'. This working definition is founded on a comprehensive approach including indi-

vidual, organizational, and technological issues and aspects.

· Chapter 3: The Topic Map Model

provides a detailed description of the basic and advanced concepts of the abstract model de-

fined by the ISO standard ISO/IEC 13250:2000 Topic Maps, the specific topic of interest

within the scope of this thesis.

· Chapter 4: Contributions of the Topic Map Model to Organizational

Memory Systems

identifies two basic contributions and points out some major limitations of the topic map

model as an underlying framework for systems supporting knowledge management efforts,

and demonstrates, how topic maps can support the evolutionary construction of organiza-

tional knowledge structures growing both in structure and extent. These findings constitute

the theoretical outcome of this thesis.

· Chapter 5: Design and Implementation of a Relational DatabaseBased Topic Map Engine

presents an approach towards the persistent storage and maintenance of large and complex

topic maps with relational database technology, as the applicative outcome of this thesis.

· Chapter 6: Conclusions and Related Work

concludes and provides an overview of major related work.

Knowledge and Knowledge Management

2.1 Knowledge

The definition of knowledge is complex, controversial and can be interpreted in different ways.

This can be mainly traced back to the fact that the term `knowledge' plays a central role in vari-

ous research disciplines (e.g. philosophy, psychology, sociology). Thus, the semantics and sig-

nificance of this term are formatively influenced by the specific interests of a particular research

discipline [Romhardt, 1998]. Within the scope of this thesis, however, the emphasis is put on

knowledge as a valuable asset for organizations -- especially for business organizations -- and

on computer-based support of its management. Therefore, a more pragmatic view of knowledge

is required.

The terms `data', `information', and `knowledge' are often used ambiguously by authors and in

daily business conversations. In order to understand the crucial issues of `Knowledge Manage-

ment', however, it is important to distinguish between them. As shown in Fig. 1, the interrela-

tionship between these terms is of hierarchical nature, which can be interpreted as follows [e.g.

Krcmar and Rehäuser, 1996; McQueen, 1998]:

· Symbols

Symbols residing on the bottom level of the hierarchy are the atomic building blocks for data.

· Data

Data consists of facts and measurements that are represented through symbols for the pur-

pose of processing and storage (usually highly structured, e.g. fields in relational database ta-

bles) without a specification of the possible use. At this level, data is relatively meaningless

to the user.

Knowledge

Information

Data

Symbols

set of symbols, e.g. "1", "9", "0",

and "."

syntax, e.g. 0.91

data in context, e.g.

exchange rate 1 = $ 0.91

integration, e.g.

rules of the exchange market

Fig. 1: The Interrelationship between Symbols, Data, Information, and Knowledge

(Source: [Krcmar and Rehäuser, 1996])

· Information

When data is placed within a meaningful context, it becomes information. Information can

therefore be regarded as data put in a specific problem context. As a short rule:

data + context = information

· Knowledge

Knowledge results, if information is combined with experience, context, interpretation, and

reflections and can therefore be regarded as a high-value form of information that is ready to

apply to decisions and actions [Davenport et al., 1998]. Unlike data and information, how-

ever, knowledge is always bound to persons and is the whole body of cognition and skills in-

dividuals make use of to resolve problems, including both theories and practical rules and in-

structions for action [Probst et al., 2000]. As a short rule:

information + experience =

knowledge

2.1.1 Types of Knowledge

Knowledge -- one of the central terms of interest within the scope of this thesis -- can be fur-

ther distinguished:

· Implicit vs. Explicit Knowledge

The most significant distinction with regard to one of the main issues of the management of

knowledge (see next section) is made between implicit (also tacit) and explicit knowledge

[Nonaka and Takeuchi, 1995]: Implicit knowledge is personal knowledge that is embedded

in individual experience and is therefore not directly accessible to others. The most common

way to share implicit knowledge is by means of highly interactive communication, since it

can hardly be reduced to rules and recipes. In contrast, explicit knowledge is usually embod-

ied in published sources like research reports, conference proceedings, text books, journal ar-

ticles, and the like. Thus, it can be more easily stored and communicated as well as shared

among persons.

· Technical vs. Cognitive Knowledge

Implicit knowledge occurs in two dimensions [Krcmar and Rehäuser, 1996]: technical and

cognitive. The former covers the range of abilities, talents, and wisdom generally understood

by the term `know-how', whereas the latter involves intangible factors such as personal be-

lief, perspective, vision, and the value system.

· General vs. (Domain-)Specific Knowledge

General knowledge is broad, usually publicly available, and independent of particular events.

Specific knowledge, in contrast, is context- or domain-specific [Zack, 1999].

· Declarative, Procedural, and Causal (or Analytical) Knowledge

Knowledge about something is referred to as declarative knowledge. Procedural knowledge

is knowledge of how something is performed and knowledge dealing with why something

occurs is called causal (or analytical) knowledge [Zack, 1999].

In the case of a law firm, for example, declarative knowledge is `knowledge of the law', namely,

the legal principles contained in statutes, court opinions, and other sources of primary legal au-

thority. Procedural knowledge is knowledge of the mechanics of complying with the law's re-

quirements in a particular situation, and causal knowledge results from analyzing declarative

knowledge as it applies to a particular fact setting (e.g. conclusions reached about the course of

action a particular client should follow in a particular situation). Whereas most of the declarative

and procedural knowledge is explicitly available, what actually makes up a successful lawyer, is

his implicit knowledge. [Edwards and Mahling, 1997]

2.2 Knowledge

Management

`Knowledge Management (KM)' is an ambiguous, often inconsistently used, and consequently

quite `fuzzy' term that refers to a broad category of business practices and/or related information

technologies that may be associated with the `handling' of organizational knowledge in general.

According to The German Research Center for Artificial Intelligence (DFKI), it is the goal of

KM to make individual (tacit) knowledge permanently available and to use already existing

knowledge as optimal as possible within an organization. Thus, KM captures and structures,

maintains, and provides enterprise-critical knowledge, information, and data that may be

unstructured, semi-formal, or formal.

The Fraunhofer IPK - Division Corporate Management defines KM as a continuous process to

increase the intellectual capital of an organization. This process is triggered by identifying those

factors that are relevant to success and by defining appropriate knowledge goals, and is based on

technological, organizational, and cultural aspects.

[Schüppel, 1996] considers KM to contain both human-centric and technologically oriented is-

sues and measures in order to optimize the creation, reproduction, distribution, and utilization of

see

http://www.dfki.uni-kl.de/km/objective.html

see

http://www-plt.ipk.fhg.de/ccwm/indexenglish/indexEng.htm

corporate knowledge. Special attention has to be directed to the mobilization of individual and

collective stocks of knowledge as well as to the learning processes for the modification and im-

provement of corporate knowledge potentials.

Within the scope of this thesis, the following working definition for KM is proposed:

`Knowledge Management (KM)' is an interdisciplinary field of research. It is the ba-

sic task of KM to provide a long-term infrastructure -- an organizational (or corpo-

rate) memory (OM/CM) -- that supports the identification, acquisition, development,

dissemination, (re-)use, and preservation of an organization's -- especially knowl-

edge workers' -- knowledge. From a non-technological point of view the emphasis is

on individual, social, organizational, and cultural aspects and prerequisites facilitat-

ing a `knowing' organization, whereas from a technological point of view the em-

phasis is on how to optimally support KM efforts with information and communica-

tion technology (ICT), mainly by means of organizational memory (or knowledge

management) systems (OMSs/KMSs). As one of the main differences to Information

Management (IM), KM operates at a more abstract level and also deals with infor-

mation that is not (yet) codified and that is highly context-dependent.

The main sources and foundations of this working definition are discussed in detail in the sub-

sections below.

2.2.1 An Interdisciplinary Field of Research

There is a wide range of research disciplines that deal with the creation, administration, mainte-

nance, and evaluation of knowledge. Although there are many perspective overlappings, each of

these disciplines has its own specific view on the topic with different research objects, research

objectives, and methods at different levels of abstraction [Frank and Schauer, 1999] (Fig. 2):

· Sociology of Knowledge

A branch of sociology studying the social processes involved in the production of knowl-

edge, especially concerned with the effects of knowledge and the social processes that might

condition either the form or content of knowledge.

· Cognitive Psychology

Cognitive psychology is concerned with mental processes and their effects on human behav-

ior and focuses on phenomena such as sensation, perception, attention, memory, learning,

language, reasoning, problem solving, and decision making. It is an empirical science that

depends on careful experimental procedures and paradigms to test theories about these men-

tal processes.

· Organization and Management Theory

From a knowledge perspective, organization and management theory focuses on individual,

collective, cultural, and organizational forms of learning, knowing, and memory as well as

the underlying organizational structures, processes, and management concepts. In this con-

nection, special attention is given to sources and keepers of organizational knowledge,

knowledge workers, and means to prevent organizational knowledge.

· Education Science

In general, education science is concerned with the transmission of knowledge and under-

standing and the development of the individual personality, for example, by teaching. Special

emphasis is put on the design and use of computer-based tutorial systems and `knowledge

media'.

· Information and Library Science

Library science is concerned with the organized archiving of huge amounts of explicit

knowledge and the efficient access to it by means of retrieval systems. Its special attention is

directed to the use of modern ICT that will gradually turn `traditional' libraries from storage

places for books into electronic knowledge centres (`digital libraries').

· Artificial Intelligence

AI is a branch of science -- mainly, but not exclusively, of computer science -- that is con-

cerned with making computers `think'. Two areas of special interest of AI are the formal rep-

resentation of knowledge (knowledge put into a form that a computer can internally store,

manipulate, and finally use to solve problems and draw inferences within the domain in

which it exists), and expert systems (computer programs that emulate the behaviour of hu-

man experts in a well-specified, narrowly defined domain of knowledge).

· Business Informatics

Business Informatics places particular importance in information systems supporting the

computer-based management of knowledge assets within organizations as well as the eco-

nomical use of such systems, considering both social and technological aspects.

Implicit Knowledge

Explicit Knowledge

Knowledge Ontology

Public

Knowledge Creation

('Learning')

Private

Knowledge as Corporate

Asset

Pragmatics of Knowledge

Utilization and

Presentation of

Knowledge

Formal Representation of

Knowledge

Artificial Intelligence

Business Informatics

Cognitive Psychology

EDS

Education Science

SOK

SOK, OMT

SOK

SOK, OMT, CP

SOK, OMT

SOK, OMT, CP, EDS

SOK, OMT, EDS

SOK, OMT, EDS, ILS

BI, OMT, CP, EDS

BI, OMT, EDS

BI, SOK, OMT, EDS, ILS

BI, AI, CP, EDS

BI, AI, EDS

BI, AI, EDS, ILS

BI, AI, CP

BI, AI

BI, AI, ILS

Infomation and

Communication

Technology

BI, AI

ILS

Information and Library Science

OMT

Organization and Management

Theory

SOK

Sociology of Knowledge

Fig. 2: Focus of Research of Various Research Disciplines on the Management of Knowledge

(Source: [Frank and Schauer, 1999])

2.2.2 Knowledge

Work

In the 1970s [Rittel and Webber, 1973] characterized what they called `wicked problems'. In

contrast to `tame problems', for which a traditional linear problem solving process is sufficient to

produce a workable solution within an acceptable time frame, wicked problems differ from tame

problems by the following criteria [Buckingham Shum, 1997]:

· A wicked problem cannot be easily defined in a way that all `stakeholders' -- people who

care about or have something at stake in how the problem is resolved -- agree on the way it

is resolved. The problem is often not clearly understood until a solution has been developed

due to interlocking issues and constraints. Therefore, wicked problems rather have better or

worse solutions than right or wrong ones.

· Most wicked problems involve several stakeholders, such as project members, members of

the management, people working on related problems, people with oversight functions, or

even externals. Many are involved in defining the problem and may add constraints to the so-

lution. Therefore, solving a wicked problem is fundamentally a social process. It requires

complex judgements about the level of abstraction at which to define the problem and there

may exist strong moral, political, or professional pressure against failures.

· Usually, the problem space is not stable. Goals, constraints, and stopping rules may change

over time, thus making it very difficult to define an objective measure of success.

· Since there is no definitive problem, there are no given alternative solutions. The whole

problem solution process may require trial-and-error iterations and ends, as soon as it runs

out of a certain resource (i.e. time, money, or `energy') unless clear stopping rules are de-

fined.

Nowadays, in the context of business organizations, the term `knowledge work' is used to de-

scribe intellectual activities that are performed by people upon data, information, and knowledge

in order to discover options, decisions, and actions with respect to the mastering of wicked prob-

lems. A knowledge worker gathers data and/or information from various sources, adds value to

this information, and distributes value-added products or services to others. Knowledge work

also differs from highly structured (and mostly automated) work where the human element does

not significantly contribute to the value of the output of the process. Organizations of which a

large portion of their mission critical tasks is knowledge work are also referred to as knowledge-

intensive organizations, such as organizations having their core competencies in the domains of

software engineering, telecommunications, consulting, law services, education, and training [e.g.

Krcmar and Rehäuser, 1996].

On the other hand, knowledge workers do not exclusively spend their time on creating knowl-

edge that would not exist without their mental efforts. Examining the work patterns of knowl-

edge workers, six more or less distinct activities can be identified [Taylor, 1998]:

1. Routine work that is hard to be separated from knowledge work (e.g. formatting an article).

2. Networking, promoting, and socializing.

3. Finding the data and information needed to produce knowledge.

4. Creating what others have probably already created, especially if this takes less time than to

search, find, and evaluate the relevancy of what has been produced by others.

5. Performing `real' knowledge work: creating what has not been created before.

6. Communicating what has been produced or learned.

Point number four in the enumeration above hints at one major problem of knowledge work. In

order to efficiently (re-)use existing results of knowledge work, an organization's structures and

culture should encourage knowledge workers to take advantage of already existing knowledge.

To achieve this, it must be ensured (1) that knowledge workers are aware of the location of al-

ready existing knowledge, and (2) that they are willing to accept and actually use knowledge `in-

vented' by others.

2.2.3 `Knowing'

Organizations

Strictly speaking, organizations cannot create knowledge `on their own'. Knowledge exists and

grows in the intellectual capacity of organizations and consists of a collection of the intellectual

capacities of individuals. Therefore, the creation of organizational knowledge is a dynamic, so-

cially complex, and iterative process of individual knowledge creation, knowledge transfer be-

tween individuals, and knowledge expansion from individual to inter-organizational levels.

Socialization

Externalization

Internalization

Combination

Implicit

Explicit

Imp

lic

Explic

Fig. 3: Four Basic Processes of Knowledge Creation and Conversion

(Source: [Krcmar and Rehäuser, 1996])

A comprehensive model of how organizations dynamically create knowledge based on a recog-

nition of the synergistic relationship between implicit (tacit) and explicit knowledge, and on the

design of social processes that create new knowledge by converting tacit knowledge into explicit

one is provided by [Nonaka and Takeuchi, 1995]. According to this model, there are four basic

processes of knowledge creation and conversion (Fig. 3):

· Socialization

Socialization is the process of sharing experiences and thereby creating implicit knowledge

via shared mental models and/or technical skills. Training `on-the-job' as apprentices learn

the craft from their masters through observation, imitation, and practice is a good example of

this type of creation.

· Externalization

The process of externalization converts implicit knowledge into explicit concepts (e.g. im-

ages, metaphors, analogies, hypotheses, or models). The writing process, whereby the written

words capture and communicate the writer's thoughts and experiences, is the most com-

monly used technique of externalization.

· Combination

Combination creates new explicit knowledge by combining already existing explicit knowl-

edge and information from a number of sources.

· Internalization

If individuals embody (or internalize) explicit knowledge gained through the two other

modes of knowledge explication (externalization and combination) in the form of shared

mental models or technical `know-how', this knowledge becomes implicit knowledge of their

own.

These four basic processes of knowledge creation and conversion form a continuous `cycle' of

organizational knowledge creation. The process typically starts with individuals developing

some insight or intuition of how to improve in performing their tasks. This implicit knowledge

may be shared with others through socialization. As long as this knowledge remains implicit,

however, the organization is unable to exploit it further. Therefore, from the organization's point

of view, the externalization of tacit knowledge into explicit concepts is vital. An organization

usually has several groups or units creating explicit knowledge at different points in time, that

may be combined and reconfigured into new forms of explicit knowledge. Finally, the cycle

closes and this new explicit knowledge is re-experienced and re-internalized by individuals as

new implicit knowledge. [Choo, 1996]

individual in a team

individual in a dyad

individual alone

Experimenter

Teller

Teacher

Listener

Learner

Any of the five roles

Creation

Transfer, Possible Creation

Creation

Possible Creation, Possible

Transfer

Possible Creation

Transfer

Possible Creation, Possible

Transfer

Creation

Possible Creation

Possible Creation, Possible

Transfer

Implicit (Tacit) Knowledge

Explicit Knowledge

Roles

Fig. 4: Roles of Individual Knowledge Creation and Transfer

(Source: [Prasser and Raven, 1996])

Another model that corresponds well to and completes the model of [Nonaka and Takeu-

chi, 1995] is provided by [Prasser and Raven, 1996]. They focus on the creation and transfer of

both explicit and implicit knowledge by individuals within three social environments (individual

alone, as part of a dyad, or as part of a team) and across a total of five roles (compare Fig. 4). In

the role of an `experimenter' an individual alone can create tacit knowledge by trying out differ-

ent ways of performing a task, and explicit knowledge may be created as a by-product of ex-

perimentation. Considering the creation of tacit and explicit knowledge by two individuals in a

dyad and its transfer between them, there are four possible roles: The `teller' acts as a transmitter

of explicit knowledge by articulating this knowledge in explicit form to the other individual in

the dyad. The `listener', in turn, is able to internalize this knowledge. If tacit knowledge has to be

transferred between two individuals in a dyad, the `teller' becomes a `teacher' and makes sure

that the `learner' acquires this knowledge through experience. As a result of these teaching ef-

forts, the `teacher' is likely to create and transfer new tacit and explicit knowledge. Within a

team, each individual can take any of the five roles described above.

2.2.4 Organizational

Memory

The term `memory' is generally understood as a system capable of storing observations and per-

ceptions beyond the duration of actual occurrences, and of retrieving them at a later point of time

within a specific context. Therefore, learning is not possible without memories. In the context of

KM, an OM is basically proposed as a prerequisite for organized organizational learning. How-

ever, an OM should not be considered analogous to a `brain' accessible by organizations. In a

generic sense, the term is simply meant to imply that an organization's knowledge workers, writ-

ten records and documents, or even data `contain' knowledge that is readily somehow accessible

with or without technological support of any kind (e.g. search and recall processes as carried out

by telephone surveys, or the technique of `brainstorming' during meetings). [Lehner and

Maier, 2000]

In research literature, OM has received a great deal of attention during the past 10 years, but

there is still no clear or agreed upon definition [e.g. Ackerman and Halverson, 1999]. The work-

ing definition for KM proposed at the beginning of this chapter is founded on two approaches

towards OM as described below.

Organizational Memory According to Pautzke

According to [Pautzke, 1989] an OM represents the collection of knowledge that is accessible to

an organization (`horizontal layers'), or -- in a wider sense -- that is at the disposal of an or-

ganization's members (`vertical layers').

In detail, there are four horizontal layers [e.g. Lehner and Maier, 2000; Romhardt, 1998]:

· Layer 1: Knowledge Shared by all Members of an Organization

This layer consists of a common language (e.g. stories, anecdotes, myths, rituals, and cere-

monies) as well as of a historically grown system of common values and norms manifested

as negotiation and behavior guidelines, manners of thinking, and paradigms. This knowledge

expressed by commonly shared values, visions, written or unwritten rules, etc. can also be

thought of as an organization's `cultural identity'.

· Layer 2: Accessible Individual and Collective Knowledge

Layer 2 comprises the individual knowledge of an organization's members that is made col-

lectively available throughout the organization. Basically, there exist three ways of how indi-

viduals can make their knowledge accessible to the organization: (1) by participating in deci-

sion making processes, (2) by participating in information sharing processes (e.g. group dis-

cussions), and (3) by explicating their knowledge by means of standard procedures, systems,

rules, etc. In contrast to the two layers described below, the current accessible content of an

organization's OM is the knowledge of both layer 1 and layer 2.

· Layer 3: Non-Accessible Individual and Collective Knowledge

Layer 3 comprises individual knowledge that cannot be used by an organization because it is

not accessible to it. Besides mostly irrelevant knowledge from an organization's members'

private spheres (e.g. leisure activities), there may also exist knowledge that is valuable to the

organization, but intentionally kept by individuals for several reasons (e.g. weakening of an

individual's own position).

· Layer 4: Meta-Knowledge of the Environment

Finally, this layer in Pautzke's model includes knowledge that generally lies outside the

realm an organization is able to access. However, an organization might have some kind of

meta-knowledge about it and might be capable -- with a certain amount of effort -- to ac-

quire this knowledge, too (e.g. by ordering a market analysis from a market research insti-

tute).

Organizational Memory According to Walsh and Ungson

According to [Walsh and Ungson, 1991] an OM is composed of several components and con-

tains at least one retention facility for gathering and searching knowledge. Its general purpose is

to connect past and present decision-making situations throughout an entire organization. Walsh

and Ungson postulate the existence of five storage components that -- together with external

sources of knowledge --make up the structure of an OM [e.g. Lehner and Maier, 2000]:

· Individuals

An organization's knowledge workers have their own memories concerning tasks, activities,

and events in and around their domain of work. Thereby, experience and observations play

an important role, and implicit knowledge is stored in forms such as principles, assumptions,

values, etc.

· Culture

Culture is defined as a learned way of perceiving, thinking, and feeling about problems that

is transmitted to members of an organization. Past experience manifests itself in the culture

of an organization (e.g. by language, common conceptions, symbols, stories, myths, and ru-

mors) and is of importance for current decisions. Thus, culture plays the role of a kind of

storage mechanism and can therefore be understood as part of the collectivization of knowl-

edge.

· Transformations

Knowledge is also embedded in numerous processes of an organization, for example, knowl-

edge about data processing (i.e. how a certain input is transformed into a specific output), or

standardized procedures. The past directs current processes and decisions, and administrative

systems serve as a mechanism that embody and preserve the knowledge embedded.

· Structure

Organizational knowledge is established in the form of social roles determining behavior

with respect to social expectations. A role conditions and limits the freedom of individual

behavior in favour of predictable behavior and safe interaction (conformity by fulfilling re-

ciprocal and established expectations). Patterns are developed out of roles and influence dis-

tribution of work as well as the entire structure of an organization. Thus, knowledge is also

coded in roles and in the resulting structure that influences decision making in an organiza-

tion.

· Ecology

The behavior of knowledge workers can also be influenced by the physical composition and

arrangement of their workplaces and environments. In large offices, for example, it is more

difficult to make `friends' and socialize knowledge.

· External Archives

When knowledge is lost or cannot be recovered when needed, it can sometimes be obtained

from external sources, for example, from former members of the organization, consultants,

government agencies, the media, etc.

2.2.5 Organizational Point of View

Traditionally, organizations have addressed KM from either a(n) organizational/management or

a technological (see next section) point of view. Managers regard an organization's knowledge

as one of its most valuable assets and are concerned with the effective use of this knowledge as

well as its qualitative and quantitative adaptation in a changing environment [Abecker et

al., 1998]. Today, however, there is a broad agreement in KM research literature that effective

and successful KM requires a hybrid solution involving both organizational and technological is-

sues and aspects [e.g. Ambrose et al., 1998; Davenport, 1996].

From an organizational point of view, knowledge has to be explicitly managed just as any other

vital asset. In a wider sense, management includes all the ways in which a particular asset or

process is handled, including, but going well beyond the work of a manager. As an analogy, fi-

nancial management includes the work of management and some specialists (i.e. accountants),

and also the work carried out by other people throughout the organization. Together they manage

the financial resources based on particular processes, structures, and supporting technology. In

the same way, KM is a process involving people throughout the organization as well as man-

agement and even key professionals, such as `knowledge officers' or `knowledge brokers'. The

main difference to the management of `traditional' assets is that the management of organiza-

tional knowledge cannot be handled by a single organizational unit (e.g. a department). KM must

be rather seen as a continuing and evolving process permeating the entire organization. From this

point of view, it requires flexible functions, structures, and systems in order to meet an organiza-

tion's changing needs as the organization itself and its internal and external environments

change.

[Hayduk, 1998], for example, points out that the creation of a knowledge-sharing and learning

organizational environment requires structures, processes, and systems of KM to be integrated

into existing structures, systems, and work processes, and individuals to feel encouraged to share

knowledge with others and to adapt shared knowledge under new conditions. The latter can be

achieved by altering traditional corporate performance measures to recognize that time spent on

knowledge-sharing activities is a legitimate and business-enhancing activity, and by incentive

systems motivating individuals to contribute to the growth of an OM.

In order to guide and help organizations in implementing KM successfully, a number of indi-

viduals and companies have developed frameworks for KM. Typically, such frameworks are

based on a generic phased model, define the specific goals and activities for each of the phases,

recommend `best practices' and methods to achieve those goals, and refer to tools that promise

to achieve those goals more efficiently. [Rubenstein-Montano et al., 2001] distinguish between

prescriptive and descriptive KM frameworks, and those that are combinations of both. Prescrip-

tive frameworks provide directions on the types of KM procedures without providing specific

details of how these procedures should be accomplished. In contrast, descriptive frameworks

characterize or describe KM and identify attributes and characteristics of KM with respect to

their influence on the success or failure of KM initiatives. The majority of KM frameworks pre-

sented in research literature are prescriptive ones.

An example of a prescriptive KM framework is the one proposed by [Probst et al., 2000].

The Knowledge Management Framework According to Probst et al.

The KM framework according to [Probst et al., 2000] distinguishes between two levels of KM

(compare Fig. 5). At the strategic level, knowledge goals are to be defined in order to control

KM at the operational level and to enable the measurement and effective adjustment of long-

term KM activities (`knowledge assessment').

At the operational level, [Probst et al., 2000] propose six interrelated `building blocks' of KM:

· Knowledge Identification

Many organizations know little about their internal skills, experts, and `knowledge networks'

because organizational changes (e.g. decentralization, restructuring, employee fluctuation)

have decreased the internal transparency of many organizations. Therefore, effective KM

must ensure that an organization can maintain a general picture of both available internal and

external knowledge in order to avoid uninformed decisions and inefficiency of knowledge

workers (`knowledge transparency'). Practices and tools supporting the identification of cor-

porate knowledge are `knowledge topographies' (identifying people possessing particular

skills and knowledge at particular levels), maps (or matrices) of knowledge assets (showing

where and how particular knowledge assets are stored), `knowledge source maps' (showing

which persons in a team, an organization, or the external environment can contribute impor-

tant knowledge to particular tasks), and geographical information systems (showing the geo-

graphical organization of knowledge assets).

Knowledge Goals

Knowledge

Assessment

Knowledge

Identification

Knowledge

Retention

Knowledge

Development

Knowledge

Sharing/

Distribution

Knowledge

Utilization

Knowledge

Acquisition

Feedback

strategic level

operational level

Fig. 5: Building Blocks of Knowledge Management

(Source: [Probst et al., 2000])

· Knowledge Acquisition

Organizations often import a substantial part of their knowledge from external sources. Sys-

tematic KM has to take into account, that knowledge can be acquired on many `knowledge

markets' to fill internal knowledge gaps (e.g. labour market, consultancy market, strategic al-

liances, pilot projects with key customers, or knowledge media).

· Knowledge Development

This building block completes knowledge acquisition. It includes all management efforts that

are consciously aimed at producing knowledge that is not yet present within the organization,

or that does not yet exist inside or outside it. The focus is on generating new skills, new

products, better ideas, and more efficient processes. Methods of collective knowledge devel-

opment are, for example, `think tanks' (groups concentrating critical organizational knowl-

edge and being entrusted with its development), `learning areas' (`germ cells' of learning and

knowledge development overlapping structures and processes of an organization without re-

placing them), and `lessons learned' (the explication, systematic collation, and making avail-

able of experiences made in projects that are likely to be of interest to future project teams

confronted with similar problems).

· Knowledge Sharing and Distribution

The sharing and distribution of knowledge within an organization are an essential precondi-

tion for turning isolated knowledge into knowledge that can be used throughout an organiza-

tion. Ongoing organizational trends towards group work, strategic alliances between organi-

zations, and `virtual' organizations make this building block a matter of priority. The two key

questions of this building block are: (1) Who should know how much about what (or be able

to do what) and to what level? (2) How can an organization facilitate the sharing and

distribution of knowledge? Since it is not necessary for everybody in an organization to know

everything, it is one of the key issues to plan and manage the infrastructure that enables the

sharing and distribution of knowledge across individual and cultural levels and conflicting

interests from an economic point of view.

· Knowledge Utilization

It is the overall aim and objective of KM to make sure that the knowledge present in an or-

ganization is applied productively. However, there is no guarantee that the knowledge ac-

quired, developed, and accumulated is actually used by knowledge workers without further

interventions from KM. Therefore, KM must establish and maintain appropriate structures

for group and individual work environments that ensure or increase the utilization of organ-

izational knowledge. Typical barriers to using and sharing knowledge are the fears of indi-

viduals of revealing their own weaknesses, or a general mistrust of `knowledge from out-

side'.

· Knowledge Retention

Knowledge -- once acquired and stored in the OM -- is not automatically available for all

the time. Basically, there are two kinds of `organizational forgetting': On the one hand, parts

of the OM are irretrievably lost to the organization (e.g. employees leave, established teams

break up, or data are destroyed), and on the other hand, access to parts of the OM is blocked

for either a time or permanently (e.g. unwillingness of individuals to pass knowledge on to

others). Therefore, the main tasks of knowledge retention (or preservation) are the identifica-

tion of key employees, attempts to bind them to the organization (e.g. by a system of incen-

tives), the improvement of socialization of knowledge between individuals or groups, and the

externalization of implicit knowledge for the purpose of easier storage in and retrieval from

the OM.

2.2.6 Technological Point of View

There has been a lively debate about the role that ICT can play for KM. On the one hand, ICT is

heavily and pervasively used in organizations, and thus qualifies as an accepted medium for the

flow of information and knowledge. On the other hand, leading KM theorists have warned

against ICT-driven KM strategies that may result in OMSs objectifying and calcifying explicit

knowledge into static and inert information disregarding altogether the role of tacit knowledge.

Although KM can perform its function without involving any form of ICT [e.g. Skok, 1999],

there is a broad agreement by now that ICT is both an integral and enabling part of KM, pro-

vided that two factors are considered [Borghoff and Pareschi, 1997]: (1) the awareness of the

limits of ICT, and (2) the fact that any application of ICT for KM will not achieve much, if it is

not accompanied by organizational, cultural, and individual aspects and issues.

Based on their experiences gained through several case studies, prototype developments, and

evaluations [Abecker et al., 1998] identify the following crucial requirements for successful ap-

proaches towards OMSs in industrial practice:

· Collection and Systematic Organization of Knowledge from Various Sources

Knowledge needed and/or produced by knowledge workers during their work processes is

usually scattered among various kinds and sources, such as paper and electronic documents,

databases, e-mails, CAD-drawings, and also tacit knowledge `stored' in the heads and minds

of individuals. The primary requirement of an OMS is therefore to enhance the accessibility

of all kinds of corporate knowledge by means of a centralized and well-structured knowledge

repository.

· Minimization of Up-Front Knowledge Engineering and Ongoing Maintenance

Even though organizations are well aware of the importance of KM, they are reluctant to in-

vest time and money into new ICT the benefits of which are distant and uncertain. Also pro-

spective users may have little time to spare for requirements definition and ongoing knowl-

edge acquisition. Ideally, an OMS exploits already available knowledge and information and

provides users with relevant content quickly. As with up-front engineering efforts, mainte-

nance should be minimal. An OMS should therefore be able to handle incomplete, potentially

incorrect, semi-structured, and frequently changing knowledge resources.

Details

Seiten
Erscheinungsform: Originalausgabe
Jahr: 2001
ISBN (eBook): 9783832450229
ISBN (Paperback): 9783838650227
DOI: 10.3239/9783832450229
Dateigröße: 1 MB
Sprache: Deutsch
Institution / Hochschule: Johannes Kepler Universität Linz – Wirtschaftsinformatik
Erscheinungsdatum: 2002 (Februar)
Note: 1,0
Schlagworte: topic maps knowledge management engine semantic networks

Autor

Knud Steiner (Autor:in)