TOWARDS LEARNING
COLLABORATIVE
NETWORKED ORGANIZATIONS
Leandro Loss, Alexandra A. Pereira-Klen, Ricardo J. Rabelo
Federal University of Santa Catarina, Department ofAutomation and Systems
GSIGMA -Intelligent Manufacturing Systems Group, BRAZIL
{loss, klen}@gsigma.uf5c.br
[email protected]{fsc.br
The concept of Learning Collaborative NeMorked Organizations merges both
the CNO and the LO paradigms. It aims at augmenting the quality of decisionmaking and of corporate governance taking inter-organizational knowledge
into account. The rationale behind the proposal is that CNOs are still lacking
research and work for enhancing their agility where rapid decision-making is
crucial for achieving their goals. In this paper knowledge management is
proposed as an approach for tackling this problem. The paper presents the first
step for a Famework for gathering information and for generating new
knowledge dynamical/v according to what users needfor given situations. The
corporate knowledge is retained, organized, shared and re-used to the benefit
of individuals and ofCNO as well as oltheir re.'pectit'e members.
1
INTRODUCTION
Emerging markets and the development of new technologies, such as Internet and
web search tools are some of the causes for breaking barriers among people, cities,
organizations, and even countries in the whole world (Friedman, 2005). Facilitators
include the easy and increasing access to communication channels and information,
as well as the production of content by and for everybody in a never-ever-thought
rate. Manuel Castells (Castells, 2006) points out that the amount of information
provided nowadays is changing life styles and the way of making business. In fact,
this reveals that in the global market there is a need to be competitive, to be aware
about the changes, to be connected to others, to collaborate, and to share knowledge.
It means that organizations should monitor their products, clients, suppliers, competitors, as well as the changes occurring in the market in order to be competitive
and, as a consequence, to survive. Therefore, it creates a kind of (dynamic) knowledge chain.
Organizations shall improve their competitiveness when they work collaboratively with each other, in a so called Collaborative Networked Organization (CNO)
paradigm. According to Camarinha-Matos (2006), a CNO is a network consisting of
a variety of entities that are largely autonomous, geographically distributed, and
heterogeneous in terms of their operating environment, culture, social capital and
goals, which collaborate to better achieve common or compatible goals, and whose
interactions are supported by computer networks.
In order to support the collaboration process, challenges such as how organizations should share common goals, build some level of trust, agree on common
practices and values as well as inter-operate based on common technological infrastructures, have risen up (Afsarmanesh, 2005). The supporting eNO structure for
Loss, L., Pereira-Klen, A.A., Rabclo, R.J., 2007, in IFIP International Federation for Information Processing,
Volume 243, Establishing the Foundation of Collaborative Networks; cds. Camarinha-Matos, L.,
Afsannanesh, H., Novais, P., Analide, c.; (Boston: Springer), pp. 243-252.
244
Establishing the foundation of collaborative networks
collaboration among organizations has been called Virtual organization Breeding
Environments (VB E). A VBE is an association of organizations and their related
supporting institutions, adhering to a long term cooperation agreement, and adoption
of common operating principles and infrastructures, targeting the growth of their
chances and their preparedness towards collaboration in potential Virtual Organizations (VOs) (Afsarmanesh, 2005). According to Rabelo (2004), VO is a dynamic,
temporary and logical aggregation of autonomous organizations that cooperate with
each other as a strategic answer to attend to a given business opportunity or to cope
with a specific need, and whose operation is achieved by a coordinated sharing of
skills, resources and information, totally enabled by computer networks.
The essential rationale of this paper is that current approaches for implementing
CNOs and their manifestations (VBEs and VOs) have not so far adequately explored
an important aspect for enhancing their agility: they have not been incorporating and
effectively using the knowledge generated along the CNO life-cycle. Additionally,
there is still a lack of adequate support for managers to consider past experiences, as
there is no - or almost none - registering and further dissemination for future use of
what has been learned individually and collectively during their life cycle. It is
argued that both good and bad experiences (CND organizational memory), which
usually comprise precious information, are simply lost during the CNOs' life time.
In this paper the concept of Learning CNOs (L-CNOs) is introduced by making
use of the Learning Organizations (LO) paradigm supported by the Knowledge
Management (KM) approach. It represents an extension of the LO concept to
strategic alliances based on the knowledge spread over the CND, where members
are usually highly heterogeneous, independent and even competitors of one another.
Learning CNO aims at augmenting the quality of decision-making and of corporate
governance taking inter-organizational knowledge into account. If a certain CNO is
able to learn with its success and failures cases, it can reduce risks and better plan
strategically its future. The objective of this paper is to introduce a systematic approach on how CNOs can become L-CNOs, making use ofKM and LO philosophies.
The content of this paper is divided as follows: section 2 stresses how learning
organizations and knowledge management concepts are combined for supporting a
Learning CNO environment; section 3 presents the importance of inherited data,
information and knowledge in this process. The supporting tools for the Learning
CNO are shown in section 4 and, finally, section 5 provides preliminary conclusions
and next steps.
2
LEARNING ORGANIZATION AND KNOWLEDGE MANAGEMENT
As time passes by, different categories of organizations have risen up with different
working styles. Figure 1 shows the main characteristics of different kinds of organizations, considering their communication scope, life cycle, decisions styles and
knowledge usage. The figure is divided in four quadrants. The lower-left quadrant
represents how traditional organizations were structured in the past: there was an
intra-organizational communication scope, focusing mostly in operational tasks.
Decisions used to be taken without considering neither past information nor knowledge, and both knowledge dissemination and human aspects are low.
Toward~
245
learning collahorative networked organizations
The upper-left and lower-right quadrants represent some forms of organizations
nowadays. Organizations presented in upper-left quadrant (CNOs) have an interorganizational communication scope, but they are mainly focused in decisions taken
to be applied in operational tasks, similar to the traditional organization view, taking
into consideration current data and information, even though they require knowledge.
In this quadrant it is possible to observe the improvement of knowledge dissemination and human aspects to an intermediate level. The organizations present in the
lower-right quadrant (LOs), however, despite they have been working in a "closed
world" due to its focus be an intra-organizational communication scope, they focus
on operational, tactic and strategic decision, they are intensive knowledge users and
human aspects are relevant.
The last quadrant, upper-right (L-CNO), is how the proposed approach sees
future organizations. L-CNOs join the inter-organizational communication scope
(CNO) - organizations willing to collaborate - with a knowledge oriented view
(LO) making highly intensive use of the knowledge in order to take decisions
focused not only in the operational level, but also in tactic and strategic decisions.
The processes are based on what has been learned and ameliorated along the time
horizon. This paper aims to contribute as a starting point towards establishing
stronger theoretical foundations of Learning CNOs.
• Communication scope:
Inter-organlzatlon
· Life cycle: creation,
operation/evolution. dlssol
• Decisions: operational,
tactical, and strategic
· Knov.1edge dissemination:
highly and Intensive
• Human aspects: high
,/
• Communication scope :
intra-<>rganization
• Life cycle : operation
• Decisions: operational
· Knowledge dissemination: low
· Human aspects: low
Communication scope:
intra -orga nlzatlon
· Ufe cycle : operation
• Decisions: operational,
tactical, and strategic
· Knowledge dissemination:
high
• Human aspects: high ..,/
~
·
Knowledge utililation
Figure 1- Organizations - past, present, and future views
Peter Senge (2004) explores Learning Organizations concept in a prescriptive
view splitting it in five disciplines that he called: systems thinking, personal
mastery, mental models, shared vision, and team learning. The first three disciplines
have particular application for the individual participant, and the last two disciplines
are applied for groups. Senge also argues that the individuals who excel in these
areas will be the natural leaders of learning organizations. From another point of
view, Nonaka (1995) coined the term of knowledge-creating in companies by the
understanding of the dynamic nature of knowledge creation and to manage such a
246
Establishing the foundation of collaborative networks
process effectively. It consists of three main elements: i) SECI, that stands for socialization, externalization, combination and internalization; ii) Ba, that can be defined
as context in which knowledge is shared, created and used through interaction; and
iii) knowledge assets, constituting the inputs, outputs, and moderating factors, of the
knowledge-creating process.
LO and KM can complement each other. On the one hand, as defined by Senge
(2004), LOs act as human beings cooperating in dynamical systems that are in a
state of continuous adaptation and improvement. As such, they need to be fed with
knowledge in order organizations can behave like that. On the other hand, KM
emerges as the supporting methodology for creating, disseminating, and promoting
knowledge use. However, the existence of KM methods and techniques is not
enough: CNO's people must be motivated to use the knowledge. In this sense, LO
can provide the support to KM activities.
2.1
Learning eNOs
CNOs, as they are nowadays, are mainly focused on daily events, specially during
the VO life-cycle. However, past data is extremely important and may be used as
source during the learning process. Data collected, stored and not used are seen as
potential knowledge. Potential knowledge is related to the knowledge that may be
extracted from the analysis of a vast amount of data. Data combination, interrogation, and speculation can lead to precious information and gives advantages for the
decision making processes in organizations (Figueiredo, 2005). Some techniques
available for this extraction are knowledge discovering in databases, intelligent data
analysis, among others.
CNOs may have the capacity to learn from their own experiences as well as to
use their data in order to become more competitive. On the one hand, in a learning
CNO, people are motivated to externalize their knowledge and the LO is responsible
for the process of knowledge generation, on the other hand KM offers support for
using such knowledge.
Once knowledge is retained and made available as formal documents, procedures
and CNO's culture, it should be used not only to improve the operational phases in
CNOs' instances (VBEs and VOs), but also in tactic and strategic planning. For
example, when a given VO is not fulfilling its deadlines, some decisions should be
taken in order to accomplish with the scheduling previously agreed (operational
phase). The reasons "why the 'that' VO was delayed" shall be analyzed and, if suitable,
VO members must receive a clearer set of instructions or the VO launching process
should be re-studiedlre-structured (strategic planning).
Considering the definition of CNO presented in the introductory section, in this
work, a Learning Collaborative Organization (L-CNO) is defined as a CNO that is
able to learn in a dynamic collaborative environment in order to continuously adapt
and improve itself. A L-CNO is able to capture CNO knowledge - stored in people's
procedures and actions or in databases - and to further organize, formalize, re-store,
and make it accessible to its actors in order to make improvements at operational,
tactic and strategic levels. This concept is supported by knowledge management and
computer systems. The kick-off process for a L-CNO is to be concerned to what has
occurred during the CNO life-cycle, it can be done via the inheritance process that is
describe in the next section.
Toward~
learning collahorative networked organizations
3
247
INHERITED DATA, INFORMATION AND KNOWLEDGE
The basis for a Learning CNO is the use of data, information and knowledge inherited
by the CNO - its organizational memory - as well as the preparedness of CNO
actors. In order to have a clear picture about inherited content, it is important to
frame the nature related to this issue. VOs have temporary and distributed behavior.
They are legally and logically dissolved after the product or service be delivered
according to the agreed contractual clauses (Rabelo, 2002).
Although CNOs manifestations) have a well defined life-cycle comprising the
creation, operation/evolution and dissolution (Camarinha-Matos, 2005), the timehorizon varies from short to long-term collaboration depending on its characteristics.
Nevertheless, CNO actors are usually able to identify what happened within their
enterprises when a VO is running, however they seldom know what happened and
the historical performance of other CNO actors. It makes the process of gathering
data, information and knowledge a hard task because they are spread over the CNO.
The task of the VO closing/dissolution, aiming to transfer the VO experience and
data, especially VO performance history, to the VBE is known as VO Inheritance
(VO-I) (Loss, 2006). This process is extremely important and demanding as such, i
is facilitated if it is supported by computer systems.
Despite the differences between VOs, VO-I approach gathers relevant data,
information and knowledge throughout the CNO and make it easily accessible, no
matter its duration. The main users of the knowledge produced by the process of
VO-I are the VBE actors. However, VOs actors can also benefit from the knowledge
base made available via VO-I.
VO Knowledge
rea lion
/
VB E Knowledge
tiliza tion
VO -I -
Figure 2 - Global Framework for VO-J
Figure 2 illustrates the overall vision of knowledge flow for VO-1. The top level
is seen as an appropriate environment for knowledge creation. VOs are the main
producers, because they work dynamically with a diversity of products and services
and they are in touch with real and unexpected situations. After the Vo dissolution
) The ones that have relationship like VBE and VOS where a VBE is the facilitator to create of VOs, or
between professional virtual communities and virtual teams.
248
Establishing the foundation of collaborative networks
phase, the knowledge produced by VOs shall be properly stored (bottom layer). The
storage process may occur either in an electronic/digital way (automatic follow-up
operations), or in a traditional way (like in printed reports and memoranda). The VBE
(middle layer) appears as the main client of the knowledge. VBE shall use knowledge in order to create value improving its preparedness as well as advertisement
campaigns to increase its reputation to the customers. The overall process occurs as
a spiral which enhances knowledge time to time flowing through the three layers.
4
SUPPORTING PROCEDURES AND TOOLS FOR L-CNOS
VO-I is more than a sum of pieces of knowledge. It provides a complete framework
with experiences, practices and case studies for CNO actors. KM is the approach
adopted in order to support the knowledge retained in the VO-I process and that will
be used in the L-CNO framework.
The KM model introduced by Nonaka (1995) was the starting point to KM in the
academia. Since then many approaches have risen up. Nissen (2000) presented an
amalgamated KM model that is used to support the proposal of this paper. This
model includes both individual and collective entities. It is arranged in six phases,
which are briefly described below:
1.
2.
3.
4.
5.
6.
Knowledge creation phase involves the discovery and the development of
new knowledge. It also includes the knowledge capture where knowledge
shall be new to a particular organization or individual.
After knowledge be created or gathered it must be organized. In order to
perform this task, knowledge representational techniques are used, like
keyword extraction, thesaurus, and ontologies to interrelate key terms and
concepts.
The formalization process involves the conversion of existing knowledge
from tacit to explicit form. It means to validate the knowledge.
Knowledge distribution phase is related to the dissemination to people and
organizations, according to the access rights previously defined.
The application phase is related to using knowledge for creating
competences.
Knowledge evolution leads in turn to further knowledge creation.
Despite its importance, KM is more a management philosophy than a mechanism. In this way, the proposed approach for L-CNO is also founded in three main
pillars: briefing and debriefing process, knowledge discovering, and knowledge
search. These pillars cover the six KM phases presented, linking it with the inheritance process.
4.1
Briefing and Debriefing
When dealing with knowledge intensive tasks, like in CNOs, it is important to map
data, information and specially knowledge, by adopting a KM practice. There are
some available practices like brainstorming, competence maps, brainwriting, heuristic
redefinition, and others. This work applies and adapts the Briefing and Debriefing
practices to the CNO context.
Towardv learning collaborative networked organizations
249
VO briefing consists in sharing with general infonnation regarding va actors.
For example, processes to be executed, management procedures concerning the
scope of specific vas, what is expected at the end of the va when it is dismantled.
In summary, it is a general guideline elaborated a priori and that describes to the
vas how to proceed and the estimated outcomes. It comprises the jormalization
and distribution (overall Figure 2), two out of the six phases described by Nissen
(2000). The former appears when the guidelines are official documents that are
made available. The latter takes place when the instructions are disseminated among
va actors in a way all va actors can have access to that knowledge.
During the process of VO debriefing va actors are required to provide feedback
regarding the occurrences during the va life-cycle which will be cross-checked
against the original plans (i.e. what was discussed in the Briefing phase). va
debriefing comprises three phases related to the KM approach presented by Nissen
(2000). The first one is related to knowledge creation (Figure 2 top-level). va actors
are motivated to communicate their mental models and visions about the va. It is
donee by sharing ideas and exchanging experiences and fonnalizing these ideas in
working plans, strategies or even rules and, as a consequence, providing the organizational learning in the eNO. Hence, they are creating tacit and explicit knowledge.
Once knowledge is written down and confinned by someone else it isjormalized. If
some adjustment is done for the briefing in a further va, knowledge evolution takes
place. For example, documents produced by past vas containing some suggestions
about a certain procedure can be used in order to compare and improve processes,
like partner search and selection.
Despite va briefing and va debriefing do not cover knowledge application
(Figure 2 middle-top-Ievels), they provide the source for applying this knowledge.
For example, information provided by the debriefing process in, its conclusions as
well as the occurrences during the va life-cycle are refined and, if suitable, used in
the briefing process in another va. This process is repeated for every va so that
va briefing and va debriefing are continuously improved. Knowledge application
(Figure 2 middle-level) is settled when the gained experience (already fonnalized
via documentation) is applied by the eNO.
It is important to emphasize that these processes require the ability to source and
integrate information into a suitable fonnat, use effective interpersonal skills to
encourage positive contributions, follow up and prepare documentation.
4.2
Knowledge Discovering in Databases
Briefing and Debriefing process deal with tacit and explicit knowledge, but not with
potential knowledge. A perspective for dealing with it is to apply techniques of knowledge discovering in databases, such as Data Mining (DM) (performed in Figure 2
bottom-level). DM is the process related to the extraction of knowledge from data
repositories aiming to identify valid, new, potentially useful and understandable
patterns (Fayyad et ai., 1996). Data provided from va actors, products, schedule,
performance indicators, trust indicators, and so forth, when collected, become potential knowledge and this kind of knowledge can be extremely important to the eNO.
The results coming from the DM process shall be shown to the managers as
transparently as possible so that they do not need to know details about implementation, data cleaning and even databases. It means that with some mouse clicks and
Establishing the foundation of collaborative networks
250
few keywords the manager will have the results of a DM algorithm showing the
patterns in data by using an easy-to-use human interface (Google-like). The functionality of knowledge discovering developed to test this concept is called Mined
Knowledge Search (MKS). It is described in section 5.
If the rules 2 generated by a DM algorithm seem to be interesting to the manager,
(s)he may use them or ask for a more detailed and intensive investigation. It allows
managers to have dynamic access, to possible solutions. The evaluation and interpretation of results is up to the manager. It is important to highlight that this process
guarantee that the results will be neither good nor accurate enough, however they
can provide some insights to the managers allowing them to take smarter and better
decisions.
4.3
Knowledge Based Information Retrieval
Five of six steps presented by Nissen (2005) were shown until now: creation,
formalization, distribution, application and evolution, but the organization (Figure 2
bottom-level) of the knowledge is still missing. Knowledge organization is as important as the other steps because there is no reason to have knowledge stored if it is not
intended to be retrieved.
Traditionally, managers have access to information by using reports based on
databases. However, they barely have an easy access to unstructured data, typically
in documents spread around the CNO, such as plain texts, e-mails, chats, or even the
reports produced during the process ofVO Briefing and VO Debriefing. It is necessary
an instrument for retrieving unstructured data in CNOs. It is done by a tool called
K-Search that is also described in section 5.
Figure 3 summarizes the
proposed approach. CNOs can
become L-CNOs when the
knowledge generated and used
during the VBENO life-cycle
is stored in non-human repositories, made available and incorporated to the CNO as routines,
systems, culture, strategies, and
so forth. The main support for
L-CNO is found in three main
pillars described in this paper briefing and debriefing, knowledge discovering in databases,
and knowledge based information retrieval. The overall approach is based in the six
aforementioned KM phases.
Figure 3 - General model of the proposed approach for
L-CNO
2 In this approach the results for OM algorithms arc presented as rules (if-then), however there are other
forms to represent it (Witten, 2005).
Towards learning collaborative networked organizations
5
251
IMPLEMENT AnON ASPECTS
The assessment of the overall approach has been done through prototypes implementation. The tool developed regarding the perspective of knowledge discovering
in databases has been called MKS. MKS is internally divided in two phases. The
first one comprises the selection of specific tables and the columns that appear to be
relevant to be used by a DM algorithm. Data used in the DM process comes from
the VBE database that usually follows a CNO reference model. Once the relevant
dataset is made available, the second phase can be started, comprising the execution
of a DM algorithm that will produce Association Rules (Witten, 2005). Association
Rules may predict any attribute and also give the freedom to predict combination of
attributes. As so many different association rules can be derived from even a tiny
dataset, the interest is restricted to those that apply to a reasonably large number of
instances and have reasonably high accuracy on the instances to which they apply to
(Witten, 2005). Phase I occurs only once, when the DM algorithm is configured.
Phase 2 is executed off-line, time to time according to a period settled by each CNO.
The results (rules) are stored in a database that is accessible via a web service. Rules
are not generated on-line because the process may take several minutes, or even
hours, and it is not practical.
An attempt to retrieve unstructured data in CNOs has been developed under the
scope of the ECOLEAD project with the K-Search functionality. It is a search
engine that allows users to perform searches with semantic embedded and in a
secure way. The K-Search functionality wraps KIM platform (Popov, 2003) which
provides infrastructure and services for automatic semantic annotation, indexing,
and retrieval of documents (knowledge organization). KIM is equipped with an
upper-level ontology and a knowledge base providing extensive coverage of entities
of general importance, but it does not cover the requirements for CNOs. An
ontology related to the CNOs area (Plisson, 2005) was plugged to KIM in order to
support documents related to CNOs and to provide shared meaning of CNO terms
into documents. Therefore the K-search functionality is able to deal with information
related to CNOs Finally, the development of a tool to help the Briefing and
Debriefing is under study.
6
CONCLUSIONS AND NEXT STEPS
This paper has introduced an approach where the concepts of knowledge management and of learning organizations are combined and applied to CNOs. Such approach
builds a learning collaborative environment by supporting a more comprehensive
decision-making, using data, information and existing knowledge about a CNO.
In order to overcome the necessity ofVBENO managers in knowing which data,
information and knowledge are right and proper, the process is split into three main
activities: in the first one there is a transformation from tacit to explicit knowledge
and vice-versa as well as the improvement of procedures by applying the processes
of VO briefing and VO debriefing. The second one is characterized by the transformation of potential knowledge that is embedded or hidden in CNOs' databases using
a data mining tool. The third one allows users to search for unstructured data supported by a tool to retrieve the already stored and semantically treated knowledge.
Establishing the foundation of collaborative networks
252
Most of the prototypes related to the presented approach are already implemented
and will be described in details in another opportunity.
The proposed approach is seen as a learning environment where the corporate
knowledge is retained, organized, shared, formalized and used to the benefit of CNO's
people, CNO's members and CNO as a whole. Next steps include the investigation
of the influence of corporate governance in learning, the final implementations as
well as examination of how the overall learning process can be used in issues related
to logistics in CNOs.
6.1
Acknowledgments
This work has been partially supported by the Brazilian councils of research and
scientific development - CNPq and CAPES. It has been developed in the scope of
the Brazilian IFM project (www.ifin.org.br) and the European 1ST FP-6 IP ECOLEAD
project (www.ecolead.org). Special thanks also to Mr. Carlos E. Gesser for his help
in part of the prototype implementation and Mr. Rui J. Tramontin Jr. for his valuable
comments.
7
I.
2.
3.
4.
5.
6.
7.
8.
9.
10.
11.
12.
13.
14.
IS.
16.
REFERENCES
Afsannanesh, H., Camarinha-Matos, L. M.; 2005. A Framework for Management of Virtual
Breeding Environments. In: PRO-VE, Proceedings. Valencia, 2005. Spain. p 35-48. Springer.
Camarinha-Matos, L. M.; Afsannancsh, H.; 2006. Collaborative Networks: Value Creation in a
Knowledge Society. In: Proceedings ofPROLAMAT'06, Shanghai, China, 14-162006.
Camarinha-Matos, L. M., Afsannanesh, H., e Ollus, M.; 2005. ECOLEAD: A holistic approach to
creation and management of dynamic virtual organizations. In: PROVE'06. Proceedings. Helsinki
2006, Finland, p 3-16. Springer.
Castells, M.; 2006. A Sociedade em Rede, volume I. Paz e Terra, 9a. edi~ao.
Fayyad, U.; Piatetsky-Shapiro, G.; Smyth P.; I 996a. From Data Mining to Knowledge Discovery:
an overview. In: Advances in Knowledge Discovery & Data Mining, pp. 1-34.
Figueiredo, S. P.; 2005. Gesmo do Conhecimento: Estraregias Competitivas para a Cria~o
e
Mobilza~
do Conhecimento na Empresa. Qualitymark, Rio de Janeiro.
Friedman, T. L.; 2005. The World Is Flat: A Brief History of the Twenty-first Century. Farrar,
Straus and Giroux.
Loss, L.; Rabelo, R. J.; Pereira-Klen, A. A.; 2006. Knowledge Management Based Approach for
Virtual Organization Inheritance In: PROVE'06. Proceedings. Helsinki 2006, Finland. p 285-294.
Springer.
Nonaka, l. e Takeuchi, H.; 1995. The Knowledge-Creating Company: How Japanese Companies
Create the Dynamics oflnnovation. New York: Oxford University Press, Inc.
Nissen, M., Kamel, M., e Sengupta, K.; 2000. Knowledge Management and Virtual Organizations,
chapter Integrated Analysis and Design of Knowledge Systems and Process.
Popov, B.; Kiryakov, A. ; Kirilov, A.; Manov, D.; Ognyanoff D.; Goranov, M.; 2003. KIM Semantic Annotation Platfonn. In: 2nd International Semantic Web Conference (ISWC2OO3).
Plisson, J. Ljubic, P.; 2005. ECOLEAD Deliverable D21.3 - Elaborating Common Ontologies.
Rabelo, R. J.; Pereira-Klen, A.; Klen, E. R.; 2004 Effective Management of Dynamic Supply
Chains, in International Journal of Networking and Virtual Organizations.
Rabelo, R. J.; Pereira-Klen, A. A.; 2002. Business Intelligence Support for Supply Chain
Management. In: BASYS02 Proceedings, Cancun 2002 p 437-444.
Senge, P. M.; 2004. A Quinta Disciplina. Best Seller, 16a. edi~ao.
Witten, l. H.; Eiber, F.; 2005 Data mining: practical machine learning tools and techniques. 2nd
Edition, Morgan Kaufmann, San Francisco.