ONTOLOGY-DRIVEN INFORMATION INTEGRATION
Networked Organisation Configuration
Alexander Smirnov, Tatiana Levashova, Nikolay Shilov
St.Petersburg Institute for Informatics and Automation of the Russian Academy of Sciences (SPIIRAS)
14 Line VO, 19178, St.Petersburg, Russia
Keywords: Networked organisation, decision support, ontology, information integration, semantic interoperability.
Abstract: Distributed networks of independent companies (networked organisations) are currently of high interest.
This new organisational form provides for flexibility, tolerance, etc. that are necessary in the current market
situation characterised by increasing competition and globalisation. Configuration of a networked
organisation is a strategic task that requires intelligent decision support and integration of various tasks
constituting the configuration problem. Achieving efficient integration of tasks is possible when it is done
taking into account semantics. The paper proposes an approach to this problem based on ontology-driven
knowledge integration. The knowledge in the approach is presented using formalism of object-oriented
constraint networks. Such formalism simplifies problem formulation and interpretation since most of the
tasks in the areas of configuration and management are constraint satisfaction tasks. The paper describes the
developed approach and the ontological model that is the core of the approach. Application of the developed
approach is demonstrated at two levels: (a) at the level of information integration within one company and
(b) at the level of information integration across a networked organisation.
1 INTRODUCTION
Global changes in the economy worldwide have led
to changes in priorities and strategies of market
players. This has caused appearance of new
network-driven organisational forms such as virtual
enterprises, extended enterprises, supply chains, etc.
A networked organization (Laudon, et. al., 2000;
Lipnack, et. al., 1994; Skyrme, 2003) is usually
defined as an organization formed by geographically
distributed independent partners on the basis of
information technologies. Efficient creation of an
effective configuration of the networked
organisation can give its members a competitive
advantage in getting an order. Hence, configuration
of the networked organisation is a problem of the
strategic level requiring intelligent decision support.
This problem has been addressed in numerous
research efforts. However most of them solve
particular tasks of the complex problem. A complex
approach is required to provide for integration of the
tasks to be solved. Among the tasks the following
most important ones can be selected (they do not
pretend to be a complete list):
1) order configuration (configuration of the
product / service in accordance with existing
constraints and customer preferences);
2) partner choice among existing companies –
potential members of the networked organization;
3) resource allocation among the networked
organisation members;
4) transportation network configuration (this
logistics related task is required due to the
distributed nature of the networked organisation);
5) configuration of technological resources of
the networked organisation members.
During the process of the networked organisation
configuration and management the above tasks have
to be solved jointly. Hence, it is reasonable to speak
about integration of them. Integration of tasks is
more than data integration. It requires integration at
the level of semantics or semantic interoperability.
In other words, it requires knowledge integration.
Knowledge management has shown its efficient
applicability in this area. It is a complex cooperative
network-centric process to support multi-object and
multi-disciplinary areas including modelling, design,
knowledge representation and acquisition, decision
124
Smirnov A., Levashova T. and Shilov N. (2006).
ONTOLOGY-DRIVEN INFORMATION INTEGRATION - Networked Organisation Configuration.
In Proceedings of the Eighth International Conference on Enterprise Information Systems - AIDSS, pages 124-131
DOI: 10.5220/0002456901240131
Copyright
c
SciTePress
support and supporting environment (Liu, et. al.,
2004). A number of efforts have been done in the
area of sharing information and processes between
applications, people and companies. However
knowledge sharing / exchange requires more than
this. It requires information coordination and
repository sharing with regard to semantics.
To address this, the paper proposes usage of the
ideas of Knowledge Logistics (KL) that stands for
acquisition, integration, and transfer of the right
knowledge from distributed sources to right persons
(decision makers) at the right time for the right
business purpose in the right context (Smirnov, et.
al., 2003a). KL with regard to individual customer
requirements, available knowledge sources, and
current situation analysis in an open information
environment addresses problems of intelligent
support of customer activities.
One of the main issues to resolve is
interoperability. It can be defined as the ability of
enterprise software and applications to interact.
Interoperability is considered to be achieved if the
interaction can, at least, take place at three levels:
data, application and business enterprise
(Interoperability Research for Networked
Enterprises Applications and Software, 2004).
Semantic interoperability assumes interaction at one
more level, namely at the level of semantics. To
provide for semantic interoperability KL uses
ontologies as one of the most advanced approaches
to knowledge mark-up and description. Ontologies
establish a joint terminology between members of a
community of interest (Semantic Web, 2005). This
makes it possible to provide for semantic
interoperability between various tasks of
configuration and management.
The paper is structured as follows. Approach
description is presented in sec. 2. Sec. 3 outlines the
principles of the central integrated ontological model
creation. The case study is given in sec. 3. Some
results are summarised in conclusions.
2 APPROACH
The KL problem in the presented approach is
considered as a configuration of a network including
end-users, knowledge resources, and a set of tools
and methods for knowledge processing located in
the network-centric environment. Such a network of
loosely coupled sources is referred to as a
knowledge source network or “KSNet” (detailed
description of the approach can be found in
Smirnov, et. al., 2003b), and the approach is called
KSNet-approach. The approach is built upon
constraint satisfaction / propagation technology for
problem solving since application of constraint
networks allows simplifying the formulation and
interpretation of real-world problems that are usually
presented as constraint satisfaction problems in such
areas as management, engineering, etc. (e.g.,
Baumgaertel, 2000).
Selected tasks are integrated using a single
integrated ontological model of the networked
organisation. In other words it can be stated that
these tasks are formulated using the semantics
provided by the ontological model. Figure 1
represents the integration of the tasks with their
input and output parameters. It also shows methods
for task solving used for the implementation of the
approach. As a notation the triad "input data –
method – output data" is used. Customer order is
considered as a driver for the entire system and the
output is a feasible configuration of the networked
organisation.
The ontology-driven architecture proposed deals
with ontologies of different types. The ontologies
are represented by means of a common notation and
a common vocabulary supported by an ontology
library. The common representation enables
performance of operations on ontology integrations
as alignment and merging, and operations on context
integrations. Main components of the ontology
library are domain, tasks & methods, and application
ontologies. All the ontologies are interrelated so that
an application ontology is a specialization both of
domain and tasks & methods ontologies.
The classification of knowledge according to the
abstraction and types (Neches, et. al., 1991)
distinguishes universal, shared, specific, and
individual knowledge abstraction levels. In the
knowledge sharing model of the system “KSNet”
(figure 2) the universal level is considered as the
common knowledge representation paradigm. The
universal level is based on the formalism of object-
oriented constraint networks represented by means
of a knowledge representation language. The
abstractions provided at this level are shared by the
ontologies stored in the ontology library. Both
shared abstraction level and specific abstraction
level are considered sharable and reusable since
ontologies of these levels share common
representation paradigm and common vocabulary.
The level of knowledge representation provides
with a common notation for knowledge description
and enables compatibility of different formats (e.g.,
KIF, OWL, etc.). Knowledge sharing level
ONTOLOGY-DRIVEN INFORMATION INTEGRATION - Networked Organisation Configuration
125
Order
Ontological model of the networked organisation
Order configuration
Customer preferences
Technological constraints
Constraint satisfaction
/
propagation
Configured order
Partner choice
Configured order
Technological capabilities,
capacities and preferences of
the potential members of the
networked organization
Coalition games
Coalition – networked
organisation members
Resource allocation
Configured order
Technological capabilities,
capacities and preferences of
the networked organization
members
Multiagent modelling
Soft computing (genetic
algorithm)
Resource allocation at resources
of the networked organization
members
Transportation network configuration
Resource allocation
Geographical locations of
resources of the networked
organization
members
Constraint satisfaction
/
propagation
Transportation routes and
schedule
Technological resource configuration
Technological resources of
networked organization
members
Constraint satisfaction
/
propagation
Production plan for the order
Feasible configuration of the networked
organisation
Figure 1: Tasks solved during configuration of the networked organisation.
Ontology library
Object-oriented constraint
networks as topic-independent
fundamental model
Knowledge representation
language
(e.g., KIF, OWL)
Domain
ontologies
Tasks &
methods
ontologies
Application
ontologies
Knowledge map
Knowledge sources including
humans
Knowledge
representation
level
Knowledge
sharing
level
Knowledge
ownership
level
Share
d
abstraction
level
Specific
abstraction
level
Universal
abstraction
level
Individual
abstraction
level
Model primitives
Language primitives
Figure 2: Ontology-driven knowledge sharing.
focuses on ontological knowledge common for a
particular area. Knowledge represented by this level
suits well for sharing and reuse, since, on the one
hand, this level does not concentrate on any specific
properties, on the other hand, knowledge of this
level is not a universal abstraction rarely taken into
account when the case considers practical
knowledge sharing and reuse. The knowledge
ownership level increases scalability of the system
regarding to number of knowledge sources that can
be attached to the system and users that can be
served.
3 ONTOLOGICAL MODEL
As a general model of ontology representation in the
system "KSNet" implementing the approach, object-
oriented constraint network paradigm (Smirnov, et.
al., 2003a) is used. This model defines the common
ontology notation used in the system. In accordance
with this representation the ontological model is
defined as follows: M = (O, Q, D, C). This
formalism includes a set of classes O and attributes
Q, Cartesian product of which is a set of variables.
Each variable may have values from a certain
ICEIS 2006 - ARTIFICIAL INTELLIGENCE AND DECISION SUPPORT SYSTEMS
126
domain D(1), …, D(n), …. The model also includes
constraints of six types: С
1
, ... , С
6
С, defining
which values the variables may take simultaneously,
and relationships between classes. To solve a
constraint satisfaction task means to assign values to
each variable so that all constraints hold. Class
"Thing" is used as a parent class for all classes of the
ontological model, i.e., any class of the ontological
model is a direct or indirect child ("is-a"
relationship) of the class "Thing".
The following types of the constraints have been
defined:
C
I
= {c
I
}, c
I
= (o, q), o O, q Q – accessory of
attributes to classes;
C
II
= {c
II
}, c
II
= (o, q, d), o O, q Q, d D
accessory of domains to attributes;
C
III
= {c
III
}, c
III
= ({o}, True False), |{o}| 2,
o O – classes compatibility (compatibility
structural constraints);
C
IV
= {c
IV
}, c
IV
= o', o'', type, o' O, o'' O,
o' o'' – hierarchical relationships (hierarchical
structural constraints) "is a" defining class taxonomy
(type=0), and "has part" / "part of" defining class
hierarchy (type=1). The most abstract class is
"Thing".
C
V
= {c
V
}, c
V
= ({o}), |{o}| 2, o O
associative relationships ("one-level" structural
constraints);
C
VI
= {c
VI
}, c
VI
= f({o}, {q}) True False,
|{o}| 0, |{q}| 0, q Q – functional constraints
referring to the names of classes and attributes.
|c| – is a number of parameters included into a
constraint (constraint cardinality).
Below, some example constraints are given:
the attribute costs (q
1
) belongs to the class order
(o
1
): c
I
1
= (o
1
, q
1
);
the attribute costs (q
1
) belonging to the class
order (o
1
) may take positive values:
c
II
1
= (o
1
, q
1
, R
+
);
instances of the class standard operation (o
2
) can
be compatible with instances of the class resource
(o
3
): c
III
1
= ({o
2
, o
3
}, True);
an instance of the class order related operation
(o
4
) can be a part of an instance of the class order
(o
1
): c
IV
1
= o
4
, o
1
, 1;
the order related operation (o
4
) is an operation
(o
5
): c
IV
1
= o
4
, o
5
, 0;
an instance of the class order related operation
(o
4
) can be connected to an instance of the class
resource (o
3
): c
V
1
= (o
2
, o
3
);
the value of the attribute cost (q
1
) of an instance
of the class order (o
1
) depends on the values of
the attribute cost (q
1
) of instances of the class
order related operation (o
4
) connected to that
instance of the class order and on the number of
such instances: c
VI
1
= f({o
4
}, {(o
4
, q
1
), (o
1
, q
1
)}).
A graphical interpretation of this model of the
networked organization at macro level is presented
in figure 3. The model contains one taxonomy
(Thing, second level classes, third level classes) and
two hierarchies (networked organization
networked organization member – Resource, and
order – order related operation).
4 CASE STUDY
This section demonstrates application of the above
approach in two areas: knowledge sharing within
one company and knowledge sharing across a
networked organisation.
Thing
Networked
organisation
Networked
organization
member
Operation
Order
Standard
operation
time
costs
time
costs
volume
Resource
Order related
operation
"is a" relationship
"part of" relationship
associative relationship
compatibility relationship
functional relationship
costs
Operation
class
attribute
First level
Second level
Third level
Figure 3: Ontological model of a networked organisation at macro level.
ONTOLOGY-DRIVEN INFORMATION INTEGRATION - Networked Organisation Configuration
127
4.1 Knowledge Sharing within One
Company
In this case study it was necessary to access
information about products and solutions stored in
various sources (documents, databases, rule bases
and Web sites) for an industrial company that has
more than 300.000 customers in 176 countries
supported by more than 50 companies worldwide
with more than 250 branch offices and authorised
agencies in further 36 countries (Hinselmann et. al.,
2004). Among the major tasks that had to be solved
the following should be outlined:
1) keep existing facilities of the applications and
avoid doubling of data;
2) extend opportunities of fast provision of
information about the company’s products by new
features (like free text search, feature prioritisation
and other);
3) provide multilingual interface;
4) implement local and Web versions of the
software;
5) index existing documents against information
stored in the databases.
To adopt the developed approach to the
company’s requirements the following tasks were
solved:
1) knowledge sources were selected and
interfaces for accessing them were developed;
2) an ontological model, which is a part of the
company’s ontology, based on available structured
data was built and extended by user-defined
elements and synonyms;
3) special methods to convert documents into
machine readable formats were developed
4) an interface to other corporate databases was
developed;
5) documents were indexed against the
ontological model vocabulary and the knowledge
map was created;
6) methods for calculation of the results
relevance, fuzzy string comparison, and document
ranking were developed.
The ontological model, which is a basis for
corporate knowledge description, was built using
structured information from databases and rule bases
(figure 4). For this purpose a number of software
modules were created that then were used for
automatic creation of the ontology. It made possible
to access all available information as if it was stored
in a single knowledge base.
As a result a system was built that based on user
(customer) requests activated appropriate knowledge
sources (Web sites, documents, databases and rule
bases) and provided access to them. The system was
successfully tested within the company.
Internet
Websites
Documents
(office, pdf)
Databases
Rule
bases
Indexing and structuring
Ontological Model:
Classes
Attributes
Constraints
Values
Knowledge Map
Experts
Manual entry
based on
Experience
Requests
analysis
Hotline
information
Etc.
User Request
Reply to the user request
Figure 4: Relation between application ontology and
knowledge sources.
4.2 Knowledge Sharing Across a
Networked Organisation
The FP6 project "Intelligent Logistics for Innovative
Product Technologies" (ILIPT) is devoted to
development of new methods and technologies to
facilitate the implementation of a new
manufacturing paradigm (Stone, et. al., 2005). This
new paradigm, "the 5-day car" will approach the
building of 'cars to order' in a reduced time scale.
ILIPT project will address the conceptual and
practical aspects of delivering cars to customers only
within several days after placing the order, the
automotive industry's exciting and radical new
business model (ILIPT, 2005). One of the tasks of
the ILIPT project is development of a common
knowledge management platform to support
interoperability within the "5-day car" production
network. This will make it possible to accumulate,
share, reuse and process knowledge across the "5-
day car" production network that in turn can
significantly help in increasing the supply chain
effectiveness and in decreasing the lead time.
This case study demonstrates creation of the
ontological model for a networked organisation (a
virtual production network). It was built using
several source ontologies and setting relationships
between their elements. Figure 5 represents
establishment of relationships between task &
ICEIS 2006 - ARTIFICIAL INTELLIGENCE AND DECISION SUPPORT SYSTEMS
128
Tasks and Methods
Ontology
Modelling
Design
Planning
Assignment
Prediction
Monitoring
Assessment
Product Unit Staff
Resource Cost Centre
Configuration
Figure 5: Task & Method and Management ontologies (a
fragment).
methods ontology and domain ontology
"Configuration".
Summing up definitions of networked
organisation configuration (e.g., Cooper et. al., 1997,
Simchi-Levi, et. al., 2000) it can be defined as
configuration of flows of products and services,
finances and information between different stages
from a supplier to a consumer / customer and
managing operational activities of procurement and
material releasing, transportation, manufacturing,
warehousing and distribution, inventory control and
management, demand and supply planning, order
processing, production planning and scheduling, and
customer service across the networked organisation.
The resulting networked organisation domain
ontology is given in figure 6. The figure presents the
class hierarchy for the classes of the taxonomy level
following the root. Management concepts are
constructed to cover various stages, functions,
decisions, and flows.
Networked organisation configuration
Networked organisation Configuration
Stages Flows
Order Functions
Figure 6: Networked organisation configuration domain
ontology: top-level classes view.
Networked organisation activities include flow of
information, materials, and finances between
different stages from suppliers to customer (figure
7). Information flow includes capacity, promotion
plans, delivery schedules, sales, orders, inventory,
quality; material flow contains row materials,
intermediate products, finished goods, material
returns, repairs, servicing, recycling, disposal;
finances flow is made up of credits, consignment,
payments (Chopra, et. al., 2001). Detailed
specializations for products and services can be
found in various product ontologies and
classification systems (e.g., UNSD, 2004, UNSPSC,
2001, NAICS, 2002) and mapped onto the presented
classification level of the material flow.
Flows
Information Funds
Material
Raw
materials
Intermediate
products
Goods
Products
Repairs
Material
returns
Servicing Recycling Disposal
Services
Figure 7: Networked organisation flows: taxonomy.
Virtual production network as a networked
organisation is a mechanism to integrate production
functions taking place at the separate stages. Most of
the functions (figure 8) happen within various
stages, some of them cross the boundaries among
several stages (Chopra, et. al., 2001). The functions
operate on the networked organisation flows.
Functions
New product
development
Marketing Operations
Distribution
Finance
Customer
service
Figure 8: Networked organisation functions: taxonomy.
Part of the built ontological model presented in
figure 9 focuses on the partner choice task. Since the
problem considered is very complex a part of the
ontology is given.
As a characteristic influencing networked
organisation performance, cost is considered. In fact
many cost items make up the total costs of the
product required by the customer and the
ONTOLOGY-DRIVEN INFORMATION INTEGRATION - Networked Organisation Configuration
129
Thing
Networked
or
g
anisation
Supplier
Configuration
Domain
Ontology
Class
BOM – Bill of Materials
BOM
Definition
Logistics
Supplier
Selection
Routing
Problem
Tasks &
Methods
Customer
Order
Stages
quantity
cost
address
material
supplied
next stage
price
discounts
quantity
response time
product variation
price
rate of innovation
service level
quantity (OP)
component
(material) (OP)
product (IP)
required quantity (IP)
component
(material)
cost (OP)
lead time
(OP)
performance
quantity (OP)
maximal cost (IP)
product (OP)
location (OP)
component
(material) (IP)
cost (OP)
[BOM Definition].[product] =
=
F
([Customer].[product variation])
[BOM Definition].[required quantity] =
=
F
([Customer].[quantity])
[Order].[component (material)] =
=
F
([BOM Definition].[product])
[Order].[quantity] =
=
F
([BOM Definition].[quantity])
cost (OP)
is-a relationship
part-of relationship
associative relationship
functional relationship
attribute
output parameter
input parameter
IP
OP
[Supplier Selection].[component (material)] = F ([BOM Definition].[component (material)]
[Supplier Selection].[maximal cost] = F ([Customer].[price])
[Supplier Selection].[location] = F ([Supplier].[address])
[Supplier Selection].[product] = F ([Supplier].[material (component) supplied])
[Supplier Selection].[quantity] = F
1
([Supplier].[next stage]) F2 ([Distributor].[availability]
[Supplier Selection].[cost] = F
1
([Supplier].[price]) F
2
([Supplier].[discounts])
Functional constraints
between supplier selection task and the domain ontology:
Figure 9: Application ontology for networked organisation management (a fragment).
production, among them manufacturing costs,
shipping costs, and other are. This means that the
complete ontological model includes all domain
ontology classes that have an influence on costs. To
simplify illustration interrelations between the
domain ontology and the set of tasks are given by
the example of the task of forming order for bill of
materials (BOM). This task defines a set of materials
and components that compose the product ordered
by the customer.
The supplier selection task follows BOM
definition task and has a set defined as input
parameters. The task also takes into account
maximal cost of the product that the customer is
ready to pay, if any. Within the limits of the
considered example the supplier selection task and
the domain ontology are interrelated by the set of
functional constraints shown in the bottom of figure
9.
Analogously, the supply chain performance
depends on supply chain configuration cost
combined with other influencing items:
[Supply Chain].[performance] =
F
1
([Supply Chain Configuration].[cost]) F
2
5 CONCLUSIONS
The paper presents an approach to semantic
information integration for intelligent decision
support in networked organisations. Usage of
ontological knowledge description made it possible
to provide for common terminology and notation
what, in turn, enabled integration of different tasks,
constituting a common complex problem.
The approach has been tested in production
related projects described in the section 4. One of
them was implemented for an industrial company
ICEIS 2006 - ARTIFICIAL INTELLIGENCE AND DECISION SUPPORT SYSTEMS
130
and was oriented to providing an access for users to
different sources containing information and
knowledge about company’s products and services.
The aim of other project was configuration of a BTO
(build-to-order) production network consisting of
several manufacturing facilities (suppliers).
Among the limitation of the approach the
complexity of the common ontological model
creation can be mentioned. However, the advantage
of the ontological model is that it is a conceptual
model of a high abstraction. Hence it can be defined
for most general concepts and detailed concepts can
be described only in the tasks.
ACKNOWLEDGEMENTS
The paper is due to the research carried out as a part
of Integrated Project FP6-IST-NMP 507592-2
"Intelligent Logistics for Innovative Product
Technologies" sponsored by European Commission,
projects supported by the Russian Academy of
Sciences # 16.2.35 of the research program
"Mathematical Modelling and Intelligent Systems"
and # 1.9 of the research program "Fundamental
Basics of Information Technologies and Computer
Systems", and project funded by grant # 05 01
00151 of the Russian Foundation for Basic
Research.
REFERENCES
Baumgaertel, H., 2000. Distributed constraint processing
for production logistics, IEEE Intelligent Systems,
Vol. 15, No. 1, pp. 40-48.
Chopra, S., Dougan, D., Taylor, G., 2001. B2B e-
Commerce Opportunities, Supply Chain Management
Review, May / June, pp. 50-58.
Chopra, S., Meindl, P., 2001. Supply Chain Management,
Strategy, Planning, and Operation, Prentice Hall.
Cooper, M.C., Lambert, D.M., Pagh, J.D., 1997. Supply
Chain Management: More Than a New Name for
Logistics, The International Journal of Logistics
Management, Vol. 8, No. 1, pp. 1-14.
Hinselmann, T., Smirnov, A., Pashkin, M., Chilov, N.,
Krizhanovsky, A. Implementation of Customer Service
Management System for Corporate Knowledge
Utilization. Proceedings of the 5
th
International
Conference on Practical Aspects of Knowledge
Management (PAKM 2004), Vienna, Austria,
December, 2004, LNAI 3336, pp. 475-486.
Intelligent Logistics for Innovative Product Technologies.
Web-site. URL: http://www.ilipt.org.
Interoperability Research for Networked Enterprises
Applications and Software, 2004. URL:
http://www.interop-noe.org.
Laudon, K. C., Laudon, J. P. 2000. Management
Information Systems: Organisation and Technology in
the Networked Enterprise. Prentice Hall International,
New York, 6
th
ed., p. 588.\
Lipnack, J., Stamps, J. 1994. The Age of the Network:
Operating Principles for the 21st century, John Wiley
& Sons, p. 264.
Liu, M., Zhong, P., Meng, X., Liu, D., Cheng, H., 2004.
Knowledge-based integrated process management in
lifecycle of product development. Concurent
Engineering: The Worldwide Engineering Grid,
proceedings of the 11
th
ISPE International Conference
on Concurrent Engineering (Sobolewsky, M., Cha, J.,
eds.), Tsighua University Press, Springer, pp. 181-186.
Neches, R., Fikes, R., Finin, T., Gruber, T., Patil, R.,
Senator, T. and Swartout, W., 1991. Enabling
technology for knowledge sharing, AI Magazine,
Vol. 12, No. 3, pp. 36-56.
Semantic Web, 2005. URL: http://www.semanticweb.org.
Simchi-Levi, D., Kaminsky, P, Simchi-Levi, E., 2000.
Designing and Managing the Supply Chain, Concepts,
Strategies, and Case Studies, McGraw-Hill, Boston.
Skyrme, D. J. 2003. The Networked Organization. URL:
http://www.skyrme.com/insights/1netorg.htm.
Smirnov, A., Pashkin, M., Chilov, N., Levashova, T,
Haritatos, F., 2003a. Knowledge source network
configuration approach to knowledge logistics,
International Journal of General Systems, Vol. 32, No.
3, pp. 251-269.
Smirnov, A., Pashkin, M., Chilov, N., Levashova, T,
2003b. Agent-based support of mass customization for
corporate knowledge management, Engineering
Applications of Artificial Intelligence, Vol. 16, No. 4,
pp. 349-364.
Stone, G., Miemczyk, J., Esser, R., 2005. Making Build to
Order a Reality: The 5-Day Car Initiative,
Strengthening Competitiveness through Production
Networks. A prospective from European ICT research
projects in the field of "Enterprise Networking", pp.
26-37.
The North American Industry Classification System
(NAICS), United States, 2002. URL:
http://www.census.gov/epcd/www/naics.html.
The UNSPSC Code (Universal Standard Products and
Services Classification Code), DAML Ontology
Library, Stanford University, 2001. URL:
http://www.ksl.stanford.edu/projects/¬DAML/-
UNSPSC.daml.
UN Classifications Registry, United Nations Statistics
Division (UNSD), 2004 February,
http://unstats.un.org/unsd/cr/registry/default.asp.
ONTOLOGY-DRIVEN INFORMATION INTEGRATION - Networked Organisation Configuration
131