Xem mẫu
- Knowledge and Process Management Volume 13 Number 1 pp 47–61 (2006)
Published online in Wiley InterScience (www.interscience.wiley.com). DOI: 10.1002/kpm.244
& Research Article
Centralized Versus Peer-to-Peer
Knowledge Management Systems
Ronald Maier* and Thomas Hadrich
¨
Department of Management Information Systems And OR, Martin-Luther-University
Halle-Wittenberg, Germany
The term knowledge management system (KMS) has been used widely to denote information
and communication technologies in support of knowledge management. However, so far
investigations about the notion of KMS, their functions and architecture as well as the differ-
ences to other types of systems remain on an abstract level. This paper reviews the literature
on KMS and distills a number of characteristics concerning the specifics of knowledge to be
managed, the platform metaphor, advanced services, KM instruments, supported processes,
participants and goals of their application. The paper then presents two ideal architectures
for KMS, a centralized and a peer-to-peer architecture, discusses their differences with the
help of two example systems and suggests that each of these architectures fits a different
type of KM initiative. Copyright # 2006 John Wiley & Sons, Ltd.
MOTIVATION (Alavi and Leidner, 2001, p. 114), and do not
answer the question whether a concrete tool or sys-
Knowledge management (KM) has been discussed tem qualifies as a KMS or, in other words, what ser-
intensively from a human-oriented and from a vices a KMS has to offer. A general frame of
technology-oriented perspective. Knowledge man- reference in the sense of a system architecture is
agement systems are seen as enabling technologies needed for the analysis of existing tools and sys-
for an effective and efficient KM. However, up- tems as well as for the development of individual
to-date the term knowledge management system KMS solutions.
(KMS) is often vaguely defined and used ambigu- Goals of this paper are to define the term KMS
ously. Examples are its use for specific KM tools, and to obtain a set of characteristics that differenti-
for KM platforms or for a combination of tools ate KMS from other types of systems (section 2), to
that are applied with KM in mind. It remains contrast two ideal architectures for KMS which are
unclear what separates KMS from other types of amalgamated on the basis of KMS architectures
systems that are also discussed as supporting KM proposed in the literature and to discuss the
initiatives. Examples are Intranet infrastructures, state-of-the-art with the help of example systems
document and content management systems, artifi- offered on the market (section 3) as well as to dis-
cial intelligence technologies, business intelligence cuss the differences between the architectures and
tools, visualization tools, Groupware or e-learning which KMS architecture fits what type of KM
systems. So far, investigations about the notion of initiative (section 4).
KMS remain on the abstract level of what a KMS
is used for, e.g. ‘a class of information systems
applied to managing organizational knowledge’ TOWARDS A DEFINITION OF
KNOWLEDGE MANAGEMENT SYSTEMS
*Correspondence to: Ronald Maier, Department of Management Even though there is considerable disagreement in
Information Systems And OR, Martin-Luther-University
Halle-Wittenberg, Germany. the literature and business practice about what
E-mail: ronald.maier@wiwi.uni-halle.de exactly KM is, there are a number of researchers
Copyright # 2006 John Wiley & Sons, Ltd.
- RESEARCH ARTICLE Knowledge and Process Management
and practitioners who stress the importance and Lewin and Minton, 1998). This definition stresses
usefulness of KMS as enabler or vehicle for the the primary goal of KMS as to increase organiza-
implementation of these approaches. A review of tional effectiveness by a systematic management
the literature on information and communication of knowledge. Thus, KMS are the technological
technologies (ICT) to support KM reveals a number part of a KM initiative that also comprises per-
of different terms in use, such as knowledge ware- son-oriented and organizational instruments tar-
house, KM software, suite, (support) system, tech- geted at improving productivity of knowledge
nology or organizational memory (information) work (Maier, 2004, p. 44ff, 55). KM initiatives can
system (e.g. Alavi and Leidner, 2001; Nedeß and be classified according to strategy in human-
Jacob, 2000; Maier, 2004, p. 79ff; McDermott, 1999, oriented, personalization initiatives and technol-
p. 104; Mentzas et al., 2001, p. 95f; Seifried and ogy-oriented codification initiatives (Hansen et al.,
Eppler, 2000; Stein and Zwass, 1995, p. 98). In addi- 1999). They can further be distinguished according
tion to these terms meaning a comprehensive plat- to scope into enterprise-specific initiatives and
form in support of KM, many authors provide initiatives that cross organizational boundaries.
more or less extensive lists of individual tools or According to organizational design, the initiative
technologies that can be used to support KM initia- can establish a central organizational unit responsi-
tives as a whole or certain processes, life cycle ble for KM or it can be a decentral initiative run by
phases or tasks thereof (e.g. Allee, 1997, p. 224f; a number of projects and/or communities. The
Binney, 2001, p. 37ff; Borghoff and Pareschi, 1998, initiative can focus on a certain type of content
p. 5f; Hoffmann, 2001, p. 78f; Jackson, 2003, p. 5f; along the knowledge life cycle e.g. ideas, experi-
Meso and Smith, 2000, p. 227ff; Ruggles, 1998, p. ences, lessons learned, approved knowledge pro-
82ff). ducts, procedures, best practices or patents.
Apart from these terms with a clear focus on KM Finally, the organizational culture of the company
or organizational memory, there is another group or organization in which the KM initiative is
of software systems that supports these approaches started, can be characterized as open, trustful, col-
called e-learning suite, learning management plat- lective where willingness to share knowledge is
form, portal, suite or system (Maier, 2004, p. 81). high or as confidential, distrustful, individual,
These platforms not only support presentation, with high barriers to knowledge sharing (see
administration and organization of teaching mate- Maier, 2004, p. 404ff for a definition of and empiri-
rial, but also interaction between and among tea- cal results about this typology of KM initiatives).
chers and students (Astleitner and Schinagl, 2000, The type of initiative determines the type of infor-
p. 114). KMS with roots in document management, mation system for its support which can be
collaboration or Groupware and learning manage- regarded as a KMS from the perspective of its
ment systems with roots in computer-based train- application environment.
ing already share a substantial portion of
functionality and seem to converge or at least be
Processes
integrated with each other. Recently, the terms
KM tools or KMS have gained wide acceptance KMS are developed to support and enhance knowl-
both in the literature and on the market. Conse- edge-intensive processes, tasks or projects (Detlor,
quently, we use the term KMS being well aware 2002, p. 200; Jennex and Olfmann, 2003, p. 214) of
that there are a number of similar conceptualiza- e.g. knowledge creation, organization, storage, retrie-
tions that complement the functionality and archi- val, transfer, refinement and packaging, (re-)use,
tectures of KMS. In the following, we will revision and feedback, also called the knowledge
summarize the most important characteristics of life cycle, ultimately to support knowledge work
KMS as can be found in the literature. (Davenport et al., 1996, p. 54). In this view, KMS pro-
vide a seamless pipeline for the flow of explicit
knowledge through a refinement process (Zack,
Goals
1999, p. 49), or a thinking forum containing interpre-
Goals are defined by the KM initiative in which the tations, half-formed judgements, ideas and other
KMS is deployed. Stein/Zwass define organiza- perishable insights that aims at sparking collabora-
tional memory information system as ‘a system tive thinking (McDermott, 1999, p. 112).
that functions to provide a means by which knowl-
edge from the past is brought to bear on present
Comprehensive platform
activities, thus resulting in increased levels of effec-
tiveness for the organization‘ (Stein and Zwass, Whereas the focus on processes can be seen as
1995, p. 95; for organizational effectiveness e.g. a user-centric approach, an IT-centric approach
48 ¨
R. Maier and T. Hadrich
- Knowledge and Process Management RESEARCH ARTICLE
provides a base system to capture and distribute services that together foster one or more KM
knowledge (Jennex and Olfmann, 2003, p. 215). instrument(s).
This platform is then used throughout the organi-
zation. In this case, the KMS is not an application
Specifics of knowledge
system targeted at a single KM initiative, but a plat-
form that can either be used as-is to support knowl- KMS are applied to managing knowledge which
edge processes or that is used as the integrating is described as ‘personalized information [ . . . ]
base system and repository on which KM applica- related to facts, procedures, concepts, interpreta-
tion systems are built. Comprehensive in this case tions, ideas, observations, and judgements’ (Alavi
means that the platform offers extensive functional- and Leidner, 2001, p. 109, 114). From the perspec-
ity for user administration, messaging, conferen- tive of KMS, knowledge is information that
cing and sharing of (documented) knowledge, i.e. is meaningfully organized, accumulated and
publication, search, retrieval and presentation. embedded in a context of creation and application.
KMS primarily leverage codified knowledge, but
also aid communication or inference used to inter-
Advanced services
pret situations and to generate activities, behaviour
KMS are described as ICT platforms on which a and solutions. Thus, on the one hand KMS might
number of integrated services are built. The pro- not appear radically different from existing IS,
cesses that have to be supported give a first indica- but help to assimilate contextualized information.
tion of the types of services that are needed. On the other hand, the role of ICT is to provide
Examples are rather basic services e.g. for colla- access to sources of knowledge and, with the help
boration, workflow management, document and of shared context, to increase the breadth of knowl-
content management, visualization, search and edge sharing between persons rather than storing
retrieval (e.g. Seifried and Eppler, 2000, p. 31ff) or knowledge itself (Alavi and Leidner, 2001, p. 111).
more advanced services e.g. profiling, personaliza- The internal context of knowledge describes the cir-
tion, text analysis, clustering and categorization to cumstances of its creation, e.g. the author(s), crea-
increase the relevance of retrieved and pushed tion date and circumstances, assumptions or
information, advanced graphical techniques for purpose of creation. The external context relates
navigation, awareness services, shared workspaces, to retrieval and application of knowledge. It cate-
(distributed) learning services as well as integra- gorizes knowledge, relates it to other knowledge,
tion of and reasoning about various (document) describes access rights, usage restrictions and cir-
sources on the basis of a shared ontology (e.g. cumstances as well as feedback from its re-use
Bair, 1998, p. 2; Borghoff and Pareschi, 1998, p. 5f; (Barry and Schamber, 1998, p. 222; Eppler, 2003,
Maier, 2004, p. 260ff). p. 125f).
KM instruments Participants
KMS are applied in a large number of application Users play the roles of active, involved participants
areas e.g. in product development, process in knowledge networks and communities fostered
improvement, project management, post-merger by KMS. This is reflected by the support of context
integration or human resource management (Tsui, in KMS. Contextualization is thus one of the key
2003, p. 21). More specifically, KMS support KM characteristics of KMS (Apitz, et al., 2002) which
instruments e.g. (1) the capture, creation and shar- provide a semantic link between explicit, codified
ing of best practices, (2) the implementation of knowledge and participants holding or seeking
experience management systems, (3) the creation knowledge in certain subject areas. Context
of corporate knowledge directories, taxonomies or enhances the simple ‘container’ metaphor of orga-
ontologies, (4) expertise locators, yellow and blue nizational knowledge by a network of artefacts and
pages as well as skill management systems, also people, of memory and of processing (Ackerman
called people-finder systems, (5) collaborative fil- and Halverson, 1998, p. 64). Communities or net-
tering and handling of interests used to connect works of knowledge workers that ‘own the knowl-
people, (6) the creation and fostering of commu- edge’ and decide what and how to share can
nities or knowledge networks, and (7) the facilita- provide important context for a KMS (McDermott,
tion of intelligent problem solving (e.g. Alavi and 1999, p. 108, 111ff). Decontextualization and recon-
Leidner, 2001, p. 114; McDermott, 1999, p. 111ff; textualization turn static knowledge objects into
Tsui, 2003, p. 7). KMS in this case offer a targeted knowledge processes (Ackerman and Halverson,
combination and integration of knowledge 1998, p. 64). Meta-knowledge in a KMS, e.g. in
Centralized Versus Peer-to-Peer Knowledge 49
- RESEARCH ARTICLE Knowledge and Process Management
the form of a set of expert profiles or the content of KMS foster the implementation of KM instruments
a skill management system, is sometimes as impor- in support of knowledge processes targeted at
tant as the original knowledge itself (Alavi and increasing organizational effectiveness.
Leidner, 2001, p. 121). The characteristics discussed above can be used as
Figure 1 gives an overview of these characteris- requirements in order to judge whether an actual
tics. The KMS is visualized by the triangle. Goals system is a KMS or not. Many systems marketed
stated by a KM initiative define the KM instru- as KMS have their foundations e.g. in document or
ments that should be supported by the KMS’s func- content management systems, artificial intelligence
tions and control their deployment. Thus, a KMS technologies, business intelligence tools, Groupware
has to be aligned with the specifics of its applica- or e-learning systems. These systems are more or less
tion environment, the types of KM initiative e.g. substantially extended with advanced services.
the strategy, scope, organizational design, type of Thus, actual implementations of ICT systems cer-
contents and cultural aspects. Participants and tainly fulfill the requirements of an ideal KMS only
communities or knowledge networks are the tar- to a certain degree. Therefore, one might imagine a
geted user groups that interact with the KMS in continuum between advanced KMS and other sys-
order to carry out knowledge tasks. The knowledge tems that can partially support KM initiatives.
tasks are organized in acquisition and deployment The characteristics discussed in this section can
processes required for the management of knowl- be seen as arguing for a certain set of services.
edge. The KMS itself consists of a comprehensive Comprehensive platform requires the inclusion of
platform rather than individual tools with infrastructure services for storage, messaging, access
advanced services built on top that explicitly and security which is built on an extensive set of
consider the specifics of knowledge as infor- data and knowledge sources. Specifics of knowledge
mation (or content) plus context. The services are call for the handling of contextualized information
combined and integrated in order to foster KM which requires integration services that describe
instruments. resources pulled together from a variety of sources.
A definition of the term KMS and a subsequent Advanced services build on top of these integration
development of architectures for KMS have to services and provide support for KM instruments.
stress these characteristics. Consequently, a KMS These knowledge services have to support the entire
is defined as a comprehensive ICT platform for col- set of acquisition and deployment processes. From an
laboration and knowledge sharing with advanced ICT perspective, these are services for publishing,
services built on top that are contextualized, inte- collaboration, learning and discovery. The knowl-
grated on the basis of a shared ontology and perso- edge services need to be tailored on the one hand
nalized for participants networked in communities. to the individual needs of participants and on the
other hand to the requirements of the roles they
perform in business processes and projects. This
calls for personalization services. Finally, participants
might need to access KMS with a host of different
appliances and applications for which access services
have to offer translations and transformation.
These services have to be aligned with each other
in architectures for KMS.
ARCHITECTURES FOR KNOWLEDGE
MANAGEMENT SYSTEMS
Architectures play an important role in MIS as
blueprints or reference models for corresponding
implementations of information systems. The
term architecture as used in MIS origins in the
scientific discipline architecture and is used in a
variety of ways e.g. application architecture, sys-
tem architecture, information system architecture
and especially software architecture. The analysis
of the definitions of KMS discussed above, of case
Figure 1 Characteristics of KMS studies of organizations using ICT in support of
50 ¨
R. Maier and T. Hadrich
- Knowledge and Process Management RESEARCH ARTICLE
KM and of KM tools and systems offered on the et al., 2002, p. 33; Zack, 1999, p. 50), market-oriented
market reveals that there are basically two ideal (e.g. Applehans et al., 1999; Bach et al., 1999, p. 69,
types of architectures of KMS: centralistic KMS Becker et al., 2002, p. 24) and several vendor-speci-
and peer-to-peer KMS. The KMS architectures sug- fic architectures (e.g. Hyperwave, Open Text Live-
gested in the following are system architectures link). The comparison of these architectures reveals
that can be used to define a framework useful (1) that each architecture suggests the establishment of
to classify individual tools and systems with a number of services organized on a number of
respect to the services they offer, (2) to analyse layers. The architectures suggest between three
which services are supported by a standard KMS and five layers that basically all follow the same
offered on the market (which is shown in this pattern in that a number of sources has to be inte-
paper) or (3) as reference architecture that helps grated so that advanced services can be built on
to design an organization-specific KMS as a combi- top. However, none of the architectures comprises
nation of tools and systems already implemented the entire set of layers needed for a KMS that fulfils
in that organization. the characteristics defined in section 2 (for a
detailed analysis see Maier, 2004, p. 250ff). For
example, Applehans et al.’s architecture has no
Centralistic architecture
integration layer with a shared taxonomy and a
Many KMS solutions implemented in organiza- repository (Applehans, et al., 1999). Bach’s architec-
tions and offered on the market are centralistic cli- ture provides the important layer of an integrated
ent-/server solutions (Maier, 2004). Figure 2 shows knowledge work place (Bach et al., 1999, p. 69).
an ideal layered architecture for KMS that repre- However, the underlying layers lack detailing.
sents an amalgamation of theory-driven (e.g. Apitz Becker et al., finally introduce the aspect of a
Figure 2 Architecture of a centralized KMS
Centralized Versus Peer-to-Peer Knowledge 51
- RESEARCH ARTICLE Knowledge and Process Management
meta-data-based integration of legacy systems into results of work on knowledge elements that has
a useful KMS (Becker et al., 2002, p. 24). However, been done offline.
the role of KMS in this architecture is reduced to a
portal. It lacks the intelligent functions that all Knowledge services
other architectures stress as being one of the key The core knowledge processes—search and retrie-
components that distinguish KMS from traditional val, publication, collaboration and learning—are
approaches. supported by knowledge services. These are key
Consequently, the ideal architecture depicted in components of the KMS architecture and provide
Figure 2 contains a superset of the services sug- intelligent functions for:
gested in the architectures mentioned above and
discovery: means search, retrieval and presenta-
is oriented towards the metaphor of a central
tion of knowledge elements and experts with
KM server that integrates all knowledge shared
the help of e.g. mining, visualization, mapping
in an organization. As in other standard architec-
and navigation tools,
tures such as the ISO/OSI model (Tanenbaum,
publication: is the joint authoring, structuring,
2003), each layer offers services to the next
contextualization and release of knowledge ele-
higher layer. The advantages are that the com-
ments supported by workflows,
plexity of the entire system is reduced and
collaboration: supports the joint creation, sharing
changes of the implementation of lower layers
and application of knowledge by knowledge
do not affect the functioning of higher layers as
providers and seekers with the help of e.g. con-
long as the interfaces of these services remain
textualized communication and coordination
the same. The arrows in Figure 2 show the data
tools, location and awareness management tools,
flow between the sources, layers and participants.
community homespaces and experience manage-
In the following, the individual layers are briefly
ment tools and
described.
learning: is supported e.g. by authoring tools and
Data and knowledge sources tools for managing courses, tutoring, learning
KMS include organization-internal sources e.g. paths and examinations.
transaction processing systems, data base systems,
data warehouses, document and content manage- Personalization services
ment systems, messaging systems and personal Main aim of personalization services is to provide a
(or group) information management systems as more effective access to the large amounts of
well as organization-external sources e.g. databases knowledge elements. Subject matter specialists or
from data supply companies, or the Internet, espe- managers of knowledge processes can organize a
cially the WWW and newsgroups. portion of the KMS contents and services for speci-
Infrastructure services fic roles or develop role-oriented push services.
The Intranet infrastructure provides basic func- Also, both, the portal and the services can be perso-
tionality for synchronous and asynchronous com- nalized with the help of e.g. interest profiles, perso-
munication, the sharing of data and documents as nal category nets and personalizable portals.
well as the management of electronic assets in Automated profiling can aid personalization of
general and of Web content in particular. In ana- functions, contents and services.
logy to data warehousing, extract, transformation
and loading tools provide access to data and Access services
knowledge sources. Inspection services (viewer) The participant accesses the organization’s KMS
are required for heterogeneous data and docu- with the help of a variety of services that translate
ment formats. and transform the contents and communication to
and from the KMS to heterogeneous applications
Integration services and appliances. The KMS has to be protected
A taxonomy or an ontology help to meaningfully against eavesdropping and unauthorized use by
organize and link knowledge elements that come tools for authentication and authorization.
from a variety of sources and are used to analyse
the semantics of the organizational knowledge
Example: Open Text Livelink 9.2
base. Integration services are needed to manage
meta-data about knowledge elements and the Open Text’s product family Livelink represents one
users that work with the KMS. Synchronization of the leading KMS platforms with a centralized
services export a portion of the knowledge work- architecture. Livelink has an installed base of over
space for work offline and (re-)integrate the 6 million users in 4500 organizations many of
52 ¨
R. Maier and T. Hadrich
- Knowledge and Process Management RESEARCH ARTICLE
Figure 3 Livelink’s components in the centralized KMS architecturey
which are large organizations.1 Figure 3 assigns knowledge sources are made available by services
Livelink’s modules to the six layers of the centra- on the infrastructure layer.
lized KMS architecture. In the following, selected
Livelink components are briefly discussed. Infrastructure services
Services called ‘activators’ extend Livelink’s search
Data and knowledge sources domain to sources like Lotus Notes data bases,
The Livelink data is stored in a relational data base Web pages (Livelink Spider), search engines and
system and the file system. Various other data and other Livelink installations (Livelink Brokered
Search). Livelink is accessed using the Intranet
1
According to Open Text Germany’s University programme infrastructure installed in an organization. The sys-
‘Knowledge management with Livelink’; see also: URL: tem’s (open) source code can be altered or
http://www.opentext.com/. The following discussion is based extended with the Livelink Software Development
on our experiences with a Livelink installation at our depart-
ment and material published by Open Text. Kit (Livelink SDK). The most common types e.g.
y
Italic descriptions refer to separate software modules that formats of office systems, can be converted to
extend Livelink’s core functionality. It depends on the actual HTML. Thus, documents can be viewed without
license agreement whether they are included or not. A variety
of additional modules can be obtained from 3rd party vendors the native application and indexed by Livelink’s
and are not considered here. search engine.
Centralized Versus Peer-to-Peer Knowledge 53
- RESEARCH ARTICLE Knowledge and Process Management
Integration services Learning services
Knowledge is stored in and represented by so- Livelink supports the design of basic courses and
called ‘‘objects’’, e.g. documents, folders, discus- question and answer tests (Livelink Learning Man-
sions or task lists that are placed in a folder hierar- agement).
chy. Meta-data is added automatically e.g.
creation/change date, creator, and manually via Personalization services
customizable categories. All meta-data are stored Livelink offers three types of workspaces that differ
in a relational data base and can be queried using mainly with respect to what groups of users are
SQL statements in so-called reports. granted privileges to access them. The enterprise
workspace is the central workspace for all users.
Discovery services A personal workspace belongs to every user with
Livelink’s full-text search engine allows basic and access restricted to this user. Project workspaces
advanced keyword searches. Additionally, the can only be accessed by participants defined by
assigned meta-data can be used for limiting the the project’s coordinator(s). The operations users
search domain. A typical search result page not and groups may perform on an object are defined
only includes a ranked list of various types of by detailed privileges at the granularity of single
objects with short descriptions e.g. documents, dis- objects. All knowledge and access services consider
cussion topics, folders or objects from further these privileges.
knowledge sources made accessible through Live- Access services
link services on the infrastructure level, but also Access to Livelink with a standard Web browser is
gives hints to what authors have been most active relatively platform-independent and not limited to
according to the actual query. Livelink’s notifica- a corporate LAN. The system can be accessed via
tion mechanism allows users to place change the Internet from every networked computer with
agents on selected folders to be notified via email a Web browser. To ease the use of the system e.g.
if changes occur. for work with a large number of documents, a cli-
ent for Microsoft Windows platforms can be
Publication services obtained optionally (Livelink Explorer). This client
Typical document management functions of Live- provides drag & drop integration into Microsoft’s
link are check-in/check-out, a versioning mechan- Windows Explorer, basic online/offline synchroni-
ism and workflows. All types of files can be zation functions and an integration into Microsoft
stored in Livelink. Optional modules provide cap- Office e.g. to check-in/check-out documents
abilities for electronic signatures (Livelink eSign), directly from Microsoft Word. If multiple installa-
functions for the management of electronic forms tions exist, the user can access them over a portal
(Livelink eForms Management), and for textual or (Livelink Unite).
graphical annotations in Adobe Acrobat’s portable
document format files (Livelink Review Manager
for Acrobat). Peer-to-peer architecture
Recently, the peer-to-peer metaphor has gained
Collaboration services
increasing attention from both, academics and
Some basic functions like discussion forums (black
practitioners (e.g. Barkai, 2001; Schoder et al.,
boards), polls, news channels, task lists and work-
2002). There have been several attempts to
flows aim at supporting collaboration. Optional
design information sharing systems or even KMS
Livelink modules offer group calendars (Livelink
to profit from the benefits of the peer-to-peer
OnTime) and electronic meetings (Livelink
metaphor (Benger 2003; Maier and Sametinger,
MeetingZone). OnTime provides a Web calendar
2004; Parameswaran et al., 2001; Susarla et al.,
with simple mechanisms to administer group 2003; ). This promises to resolve some of the short-
appointments. MeetingZone comprises a set of
comings of centralized KMS e.g.
meeting support tools integrated into Livelink e.g.
whiteboard, chat, shared desktop and objects to be to reduce the substantial costs of the design,
used during the meeting. The Livelink Skills implementation and maintenance of a centra-
Management module offers the management of lized knowledge server,
an extended set of data about users. Livelink Com- to reduce the barriers of individual knowledge
munities comprises four smaller modules (forums, workers to actively participate and share in the
blogs, FAQ and calendar) that facilitate interaction benefits of a KMS,
between participants and allows for arranging to overcome the limitations of a KMS that focuses
community workspaces. on organization-internal knowledge whereas
54 ¨
R. Maier and T. Hadrich
- Knowledge and Process Management RESEARCH ARTICLE
many knowledge processes cross organizational other, thus forming a separate peer-to-peer net-
boundaries, work. Requests from peers are always handled
to include individual messaging objects (emails, by the connected super peer and eventually for-
instant messaging objects) into the knowledge warded to other super peers. As in the assisted
workspace and architecture, a direct connection between peers is
to seamlessly integrate the shared knowledge established, once a peer with the desired resource
workspace with an individual knowledge work- is found.
er’s personal knowledge workspace. The more functionality for central coordination is
required in a peer-to-peer system, as is the case in a
However, there is no common architecture or an
KMS, the more likely it is that some kind of assis-
agreed list of functions yet for this type of KMS.
tance by a server is needed to coordinate the
Generally, the peer-to-peer label is used for differ-
system. Consequently, Figure 4 depicts the archi-
ent architectures (e.g. Dustdar et al., 2003, p. 170ff).
tecture of a peer and a server to assist the network.
Firstly, the assisted peer-to-peer architecture requires
Both architectures basically consist of the same
a central server e.g. to authenticate all users to act
layers as the architecture of centralized KMS.
as a global search index. Peers send search
Thus, in the following only the differences to the
requests to the server that directs peers to
centralized architecture are discussed.
resources which are then transferred directly
between the peers. Secondly, the pure peer-to-peer
Peer
architecture does not have any central authentica-
tion or coordination mechanism. Every peer pro-
vides complete client and server functionality Infrastructure services
(‘servents’). Lastly, the super peer architecture is in Personal data and knowledge sources are made
between assisted and pure architectures. Super accessible by extract transformation and loading
peers are peers with a fast and stable network con- services. Infrastructure services also provide the
nection. A peer is connected to one single super peer-to-peer infrastructure for locating peers,
peer, thus forming clusters of peers in the net- exchanging data with other peers and assuring
work. Super peers are also connected to each security of the personal knowledge base.
Figure 4 Architecture of server and peer
Centralized Versus Peer-to-Peer Knowledge 55
- RESEARCH ARTICLE Knowledge and Process Management
Integration services Personalization services
A personal taxonomy or an ontology are the foun- Profiles and push services ease access to the orga-
dation for definition and handling of meta-data of nized collection of (quality approved or even
the knowledge objects in the personal knowledge improved) knowledge elements that the subject
base. The knowledge base comprises private, pro- matter specialists administer.
tected and public areas. Private workspaces con-
tain information that is only accessible for the Access services
owner of the private workspace. Public work- These services are restricted to the administration
spaces hold knowledge objects that are published of the server, the central knowledge structure and
via the Internet and accessible by an undefined the profiles for personalization.
group of users. Protected workspaces contain
knowledge objects that are accessible to a single
Example: Groove Networks Groove 2.5
or a group of peers that the owner explicitly grants
access. The product Groove from Groove Networks targets
collaboration in small groups and is based on the
Knowledge services peer-to-peer metaphor. In the following, its func-
Just as in the centralized case, these services build tions are discussed briefly using the layers of the
upon the knowledge base. The main difference is peer-to-peer architecture (see Figure 5).2
that the knowledge repository now is spread across
a number of collaborating peers that have granted
Peer
access to parts of their knowledge repositories.
Data and knowledge sources
Personalization services The data resides in XML stores on the local hard
Contents and services are personalized based on disks of the peers. It is possible to import calendar
individual user profiles and on centralized perso- items, emails and contacts from MS Outlook,
nalization services provided by the server. to integrate MS Sharepoint workspaces (discus-
sions and documents are synchronized, other
Access services elements of a Sharepoint workplace are stored in
There are no differences compared to the centra- the forms tool) and to import data from MS
lized KMS architecture. Project. File viewers can be downloaded for com-
mon file types.
Server
Infrastructure services
The data store is managed by a storage service that
Infrastructure services ensures persistence of Groove’s workspaces. Local
A server might access a number of additional, data and messages to other peers are encrypted by
shared data and knowledge sources and assist the a security service. A user normally owns one
peers with additional services. The peer-to-peer account that includes one or more identities. Every
infrastructure might also provide services for look- identity has a pair of public/private keys and a
up and message handling that improve the effi- fingerprint for encryption and authentication. It is
ciency of the distributed KMS. possible to exchange text or voice messages.
Peer connection services determine IP addresses
Integration services of other peers and handle communication using
A shared taxonomy or ontology for the domain is the proprietary simple symmetrical transmission
offered which is handled e.g. by a network of sub- protocol (SSTP). Device presence services handle
ject matter specialists. This addresses the challenge the detection of other peers and their online/offline
in a totally distributed KMS that the various knowl- status. The Groove Development Kit (GDK) pro-
edge bases cannot be integrated and thus pose a vides an environment for programming software
problem for e.g. the interpretation of search results extensions using Microsoft software components
by the knowledge worker. The server might offer (COM) and programming languages like VB.NET,
replication services to peers that sometimes work Cþþ or C#.
offline.
Knowledge services 2
The following discussion is based on our experiences with a
There are no central services in addition to the Groove installation at our department, on Pitzer, 2002 and mate-
peers’ services. rial published by Groove Networks.
56 ¨
R. Maier and T. Hadrich
- Knowledge and Process Management RESEARCH ARTICLE
Figure 5 Groove’s components in the architecture of decentralized KMS
Integration services tool (task list). A sketchpad (whiteboard) and an
Knowledge workers collaborate in workspaces that outline tool (structured list) offer basic support
contain a number of tools. Every user can create a for brainstorming sessions. A group of users can
workspace, assign tools and invite other users to jointly browse Internet/Intranet-pages with co-
join. All knowledge elements like basic text, docu- browser functionality using Microsoft Internet
ments, calendar items or images are stored in this Explorer. A ‘navigate together’ option synchronizes
workspace and are only visible to the members of the interface of the workspace. Awareness services
this workspace whose privileges depend on their provide information about current activities of
role (guest, participant or manager). There is no other users, e.g. the workspace and the tools they
central taxonomy or ontology. Changes in work- currently access. Information about users is distrib-
spaces are continuously transmitted to all peers. If uted within Groove or by e-mail.
a peer goes offline, the differentials are synchro-
nized when he switches back online. Discovery and learning services
Groove clearly emphasizes collaboration functions
Publication services and lacks discovery services like a full-text search
Groove offers no advanced publication services engine as well as learning services.
except the review cycle tool for joint revision of
documents and a function that allows users to Personalization services
simultaneously co-edit MS Word and MS Power- Groove allows for simple adaptation of the user
point documents. Files can be stored in a basic hier- interface, e.g. design of skins and selection of
archical folder structure in the files tool. The Groove services offered in particular workspaces.
picture and the notes tools are for storing and view- However, there are no solutions that consider
ing pictures and text. Structured data is stored in user profiles when nvoking services on the lower
forms created with the forms tool. levels of the architecture.
Collaboration services Access services
Basic collaboration tools offered by Groove are a The workspaces are accessed by a MS
group calendar, a group contact list, a discussion Windows client called transceiver with a drag
forum, meeting minutes and a project manager and drop interface for files. The Groove explorer
Centralized Versus Peer-to-Peer Knowledge 57
- RESEARCH ARTICLE Knowledge and Process Management
offers an alternative user interface with the same DISCUSSION
functionality. Each user creates an account secured
by a password. Table 1 shows to what extent Livelink and Groove
fulfill the requirements that have been identified in
section 2 and for what type of KM initiative as
Server
defined in the requirement goals these systems are
A peer-to-peer network bears challenges with suited.
respect to central management tasks like license Livelink is a KMS that offers a comprehensive
management or coordinating resource utilization platform and functions at every level of the centra-
e.g. bandwidth or disk capacity. Groove addresses lized architecture. With roots in document manage-
them with centralized servers. ment, Livelink’s focus is on explicit knowledge,
with advanced functions for contextualization,
Data and knowledge resources publication and discovery across formats, plat-
Other systems like enterprise resource planning forms and the boundaries of a corporate LAN.
(ERP) software or customer relationship manage- Also, Livelink supports collaboration based on joint
ment systems (CRM) can be integrated by software authoring and sharing of documents. Although
agents called bots. Data needed and produced by Livelink can be used (almost) out-of-the-box as a
Groove’s server application resides in a local data basic KMS platform, most implementations adapt
store. the user interface to corporate style guides and
extend the integration and infrastructure capabil-
Infrastructure services ities to cover organization-specific data and knowl-
The server offers relay services to ensure stable and edge sources. It is certainly more ambitious to
fast communication between peers. If a peer’s con- combine and integrate Livelink’s knowledge ser-
nection to the network is slow, large files are sent to vices into KM instruments. Open Text’s offerings
and distributed by the relay server (‘fanout’ func- here are limited to a basic skill management instru-
tionality). Peers behind firewalls can communicate ment and a module to set up community spaces.
with the relay server using the Hypertext Transfer Groove can be characterized as a peer-to-peer
Protocol (HTTP). The server then transmits the collaboration tool that in its current form lacks a
data to the addressed peers using the preferred number of functions required in a KMS, but is cer-
SSTP. tainly a promising candidate for an integration of
Moreover, the server offers functions for the the missing functions e.g. discovery services like
management of licenses, distribution of software full-text search or navigation of workspaces, a tax-
updates, monitoring of Groove’s usage, directory onomy or ontology that integrates the knowledge
services for exchanging user information, a public currently scattered across multiple workspaces,
key infrastructure (PKI) and basic account manage- customizable meta-data, personalization and a
ment for using one Groove account on multiple tighter integration of the tools in a workspace e.g.
computers. Groove allows monitoring of network the review cycle and files tool.
usage, of workspaces and their tools as well as However, there are still serious technical chal-
the activity of single users. lenges that have to be overcome in peer-to-peer
computing in general. These challenges concern
Integration services connectivity e.g. locating peers that do not have
Another part of the relay services addresses the public IP addresses, security and privacy e.g. the
synchronization of peers. Messages to peers cur- risk of spreading viruses, unauthorized access to
rently offline are temporarily stored and forwarded confidential and private information and the
when peers go back online. The data resides in a installation of unwanted applications, availability
local cache. and scalability e.g. concerning searches in the
flat structure of the distributed search domain
Knowledge and personalization services (Barkai, 2001, p. 264ff). There are also organiza-
Due to the fact that the centralized server is tional issues that have to be resolved before a
designed for coordinating a peer-to-peer network peer-to-peer KMS can be fully deployed in an
and for the technical integration of legacy systems, organization e.g. the participation issue, i.e. there
it offers no such centralized services. have to be incentives to actively participate in the
peer-to-peer network in order to foster information
Access services sharing and avoid the free rider issue, the trust
The user interface for the administrator is a stan- issue, i.e. participants have to believe in the secur-
dard Web browser. ity and reliability of the peer-to-peer infrastructure
58 ¨
R. Maier and T. Hadrich
- Knowledge and Process Management RESEARCH ARTICLE
Table 1 Examples for centralized and peer-to-peer systems compared
Requirements Open Text Livelink 9.2 Groove Networks Groove 2.5
Platform Integrated set of functions for all areas Integrated set of functions with strong
required for KMS; multi-user system for emphasis on collaboration; limited number
1000þ users; easily scalable of peers, because network traffic and
management of privileges might prevent
scalability
Advanced services Advanced services for publication and Advanced services restricted to collabora-
discovery; basic support for collaboration, tion and awareness; basic support for
contextualization, integration and integration and workspace management
personalization
KM instruments Basic skill management, communities None
Processes Organize, store, search, retrieval, transfer, Store, transfer, revision, feedback
revision, feedback
Specifics of knowledge Mainly stable, documented but also ad hoc, Focus on ad hoc and co-authored knowl-
co-authored knowledge; customizable meta- edge including text and voice communica-
data for contextualization; no support for tion; no meta-data; no support for stages of
stages of knowledge knowledge
Participants More rigidly defined small to large teams Small, flexible teams, often crossing organi-
within an organizational setting zational borders
Type of initiative Open Text Livelink 9.2 Groove Networks Groove 2.5
Strategy Codification Personalization
Organizational design Central Decentral
Content Lessons learned, (approved) knowledge Individual contents, ideas, results
products, secured knowledge as well as of group sessions and experiences
ideas, experiences and individual contents
Organizational culture Both types of culture (restrictive or loose Open, trustful, collective
user privileges)
or the coordination issue, i.e. structuring (organiz- strategy to a combination and integration of func-
ing, packaging) and quality management (revision, tions for handling internal and external context,
feedback) of the knowledge contained in a peer- locating experts, skill management, etc. which
to-peer network have to be supported in order to bridges the gap to a personalization strategy (Maier
avoid information overload (Susarla et al., 2003, 2004, p. 506). Advanced functions supporting colla-
p., 133ff). boration in teams and communities, tools linking
Consequently, a centralized KMS like Livelink knowledge providers and seekers as well as e-
seems to be better suited for a KM initiative that learning functionality have been integrated into
can be described as a codification initiative many centralized KMS. KMS offered on the market
restricted to the organization’s boundaries, mana- differ with respect to the extent and intensity with
ged by a central organizational unit and fostering which they cover the services included in the cen-
the handling of all types of knowledge. A peer-to- tralized architecture. Some focus on learning man-
peer information sharing system like Groove tar- agement (e.g. Hyperwave), some on integration
gets a KM initiative that can be described as a per- (e.g. Lotus Notes/Workspace), on discovery (e.g.
sonalization initiative involving members from a Verity) publication (e.g. Livelink), collaboration
number of institutions. Thus the initiative is mana- (e.g. CommunityBuilder) or personalization and
ged decentrally requiring an open, trustful, collec- access (e.g. SAP Portals).
tive organizational culture and a focus on the
exchange of individual knowledge, ideas and
experiences. CONCLUSION
Generally, there has been a shift in perspective of
KMS vendors as well as organizations applying This paper has studied the notion of the term KMS
those systems from a focus on documents contain- and provided a definition and a set of characteris-
ing knowledge and thus from a pure codification tics of KMS. Ideal architectures for centralized and
Centralized Versus Peer-to-Peer Knowledge 59
- RESEARCH ARTICLE Knowledge and Process Management
peer-to-peer KMS have been contrasted and illu- ¨
Bach V, Vogler P, Osterle, H (eds). 1999. Business
strated with the help of two example systems. Knowledge Management: Praxiserfahrungen mit Intranet-
¨sungen. Springer: Berlin.
basierten Lo
The systems’ ability to support KM initiatives has Bair J. 1998. Dimensions of KM Technology Selection.
been discussed using the KMS characteristics. Gartner Group, Research Note T-05-0592.
Each of these systems targets a different type of Barkai D. 2001. Peer-to-peer Computing. Technologies for
KM initiative. Summing up, it seems that centra- Sharing and Collaborating on the Net. Intell Press:
lized KMS offered on the market more and more Hillsboro (OR).
Barry CL, Schamber L. 1998. Users’ criteria for relevance
live up to the expectations of organizations ready evaluation: a cross-situational comparison. Information
to apply ICT to support a KM initiative. Peer- Processing & Management 34(2–3): 219–236.
to-peer KMS promise to resolve some of the short- Becker J, Neumann S, Serries T. 2002. Integration von
comings of centralized KMS, especially concerning Workflow- und Wissensmanagement zur Flexibilisier-
the time-consuming effort to build and maintain a ung industrieller Geschaftsprozesse. Industrie Manage-
¨
ment 18(3): 23–27.
central knowledge repository, but also suffer from Benger A. 2003. Dezentrales, selbstorganisierendes Wis-
technical and organizational issues still unresolved. sensmanagement. In Wissensmanagement: Potenziale—
This is especially true for KMS that span organiza- Konzepte—Werkzeuge, Proceedings of the 4th Oldenburg
tions targeted at cooperation partners, joint ven- Conference on Knowledge Management, Gronau N (ed.).
tures and alliances. University of Oldenburg, GITO-Verlag: Berlin; 155–
170.
One of the biggest research questions still unre- Binney D. 2001. The knowledge management spec-
solved is how to design such solutions. Challenges trum—understanding the KM landscape. Journal of
in the design of KMS are on the one hand the inte- Knowledge Management 5(1): 33–42.
gration with existing applications, such as enter- Borghoff UM, Pareschi R (eds). 1998. Information
prise systems or office tools, and on the other Technology for Knowledge Management. Springer: Berlin.
Davenport TH, Jarvenpaa SL, Beers MC. 1996. Improving
hand the integration of individual and organiza- knowledge work processes. Sloan Management Review
tional knowledge bases. Another challenge is 37(4): 53–65.
which models to use for the design of KMS and Detlor B. 2002. An informational perspective towards
how to integrate the modelling efforts with busi- knowledge work: implications for knowledge manage-
ness process modelling. Some first approaches ment systems. In Knowledge Mapping and Management,
White D (ed.). IRM Press: Hershey; 195–205.
focus the definition of knowledge portals that sup- Dustdar S, Gall H, Hauswirth M. 2003. Peer-to-Peer-
port establishment of links between knowledge cre- ¨r
Architekturen. In Software-Architekturen fu Verteilte
ated in a certain task within a business process and Systeme, Dustdar S, Gall H, Hauswirth M (eds).
knowledge required in another task or that support Springer: Berlin; 161–198, URL: http://www.infosys.
a number of predefined opportunities in which tuwien. ac.at/staff/hg/SA-VS/07_P2P_pages.pdf [30
March 2005].
employees would switch from working on a busi- Eppler MJ. 2003. Managing Information Quality: Increasing
ness process into a learning situation. the Value of Information in Knowledge-intensive Products
and Processes. Springer: Berlin.
Hansen MT, Nohria N, Tierney T. 1999. What’s your
REFERENCES strategy for managing knowledge? Harvard Business
Review 77(3–4): 106–116.
Ackerman MS, Halverson C. 1998. Considering an Hoffmann I. 2001. Knowledge management tools. In
organization’s memory. In Proceedings of CSCW’98. Knowledge Management: Best Practices in Europe, Mertins
39–48, URL: http://www.eecs.umich.edu/ ackerm/ K, Heisig P, Vorbeck J (eds). Springer: Berlin; 74–
pub/98b24/cscw98.om.pdf [last access: 30 March 94.
2005]. Jackson, C. Process to Product: Creating Tools for Knowl-
Alavi M, Leidner DE. 2001. Review: knowledge manage- edge Management. URL: http://www.brint.com/
ment and knowledge management systems: concep- members/online/120205/jackson/ [last access: 30
tual foundations and research issues. MIS Quarterly March 2005]
25(1): 107–136. Jennex M, Olfmann L. 2003. Organizational memory.
Allee V. 1997. The Knowledge Evolution: Expanding In Handbook on Knowledge Management: Knowledge
Organizational Intelligence. Butterworth-Heinemann: Directions (Vol. 2), Holsapple CW (ed.). Springer:
Boston. Berlin; 207–234.
Apitz R, Lattner A, Schaffer, C. 2002. Kontextbasiertes
¨ Lewin AY, Minton JW. 1998. Determining organizational
Wissensmanagement in der Produktentwicklung als effectiveness: another look, and an agenda for research.
Grundlage fur anpassungsfahige Unternehmen. Indus-
¨ ¨ Management Science 32(5): 514–553.
trie Management 18(3): 32–35. Maier R. 2004. Knowledge Management Systems: Informa-
Applehans W, Globe A, Laugero, G. 1999. Managing Knowl- tion and Communication Technologies for Knowledge
edge: A Practical Web-Based Approach. Addison-Wesley: Management (2nd edn). Springer: Berlin.
Reading, MA. Maier R, Sametinger J. 2004. Peer-to-peer information
Astleitner H, Schinagl, W. 2000. High-level Telelernen und workspaces in infotop. International Journal of Software
Wissensmanagement—Grundpfeiler virtueller Ausbildung. Engineering and Knowledge Engineering 14(1): 79–
Frankfurt/Main. 102.
60 ¨
R. Maier and T. Hadrich
- Knowledge and Process Management RESEARCH ARTICLE
McDermott R. 1999. Why information technology Schoder D, Fischbach K, Teichmann R (eds). 2002. Peer-
inspired but cannot deliver knowledge management. ¨
to-peer. Okonomische, technologische und juristische Per-
California Management Review 41(4): 103–117. spektiven. Springer: Berlin.
Mentzas G, Apostolou D, Young R. 2001. Knowledge net- Seifried P, Eppler MJ. 2000. Evaluation fuhrender Knowl-
¨
working: a holistic solution for leveraging corporate edge Management Suites. Wissensplattformen im Ver-
knowledge. Journal of Knowledge Management 5(1): 94–106. gleich. NetAcademy Press: St. Gallen (CH).
Meso P, Smith R. 2000. A resource-based view of organi- Stein E, Zwass V. 1995. Actualizing organizational mem-
zational knowledge management systems. Journal of ory with information systems. Information Systems
Knowledge Management 4(3): 224–234. Research 6(2): 85–117.
Nedeß C, Jacob U. 2000. Das Knowledge Warehouse vor Susarla A, Liu D, Whinston AB. 2003. Peer-to-peer enter-
der Gefahr der Komplexitatsfalle. In Wettbewerbsvorteile
¨ prise knowledge management. In Handbook on Knowl-
durch Wissensmanagement. Methodik und Anwendungen edge Management: Knowledge Directions (Vol. 2),
des Knowledge Management, Krallmann H (ed.). Holsapple CW (ed.). Springer: Berlin; 129–139.
Schaffer-Poeschel: Stuttgart; 91–116.
¨ Tanenbaum AS. 2003. Computer Networks (4th edn).
Parameswaran M, Susarla A, Whinston AB. 2001. P2P Upper Saddle River, Prentice Hall PTR: New Jersey.
networking: an information sharing alternative. IEEE Tsui E. 2003. Tracking the role and evolution of commer-
Computer 34(7): 1–8. cial knowledge management software. In Handbook on
Pitzer B. 2002. Using Groove 2.0. Indianapolis, Que. Knowledge Management: Knowledge Directions (Vol. 2),
Ruggles RL. The state of the notion: knowledge manage- Holsapple CW (ed.). Springer: Berlin; 5–27.
ment in practice. California Management Review 40(3): Zack MH. 1999. Managing Codified Knowledge. Sloan
80–89. Management Review 40(4): 45–58.
Centralized Versus Peer-to-Peer Knowledge 61
nguon tai.lieu . vn
