ARTICLE IN PRESS
0268-4012/$ - se
doi:10.1016/j.iji
�CorrespondE-mail addr
International Journal of Information Management 27 (2007) 86–103
www.elsevier.com/locate/ijinfomgt
Improving the quality of collaboration requirements forinformation management through social networks analysis
Carla Sofia Pereiraa,b,�, Antonio Lucas Soaresa,c
aInstituto de Engenharia de Sistemas e Computadores do Porto, Campus da FEUP, Rua Dr. Roberto Frias 378, 4200-465 Porto, PortugalbEscola Superior de Tecnologia e Gestao de Felgueiras, Instituto Politecnico do Porto, Casa do Curral, Rua do Curral, Apartado 205,
4610-156 Felgueiras, PortugalcFaculdade de Engenharia da Universidade do Porto, Rua Dr. Roberto Frias, 4200-465 Porto, Portugal
Abstract
The right choice of the method of organizational analysis to use is a key factor in the process of requirements analysis
and specification of an information system. Although a high number of approaches of organizational analysis exist, the
choice of the most appropriate option for each concrete case will influence the quality of the results obtained in the analysis
of requirements and consequent specification. This paper presents a new way for organizational analysis to improve the
quality of the requirements of systems that support information management and where collaboration is an important
aspect. This is achieved through the application of the social network analysis approach, applied to refine, classify and
prioritize the requirements for collaboration and information management in an organization. The paper begins by
analysing shortly content management systems and wiki systems as IT platforms for collaboration and information
management. After having described the method, a practical case of application of SNetCol method to an R&D institution
is presented. The paper finishes by presenting the results of the evaluation of the two particular technological options
considered for satisfying the specified requirements are described.
r 2006 Elsevier Ltd. All rights reserved.
Keywords: Social network analysis; Analysis and specification of collaborative systems; Information management; Content management
systems; Wikis
1. Introduction
Obtaining pertinent, consistent and up-to-date information across a company is a complex process. For thisreason, corporations have been seeking to develop a number of information technology systems to assist with theinformation management of their business processes. Such systems aim to improve the way in which information isgathered, managed, distributed, and presented to people in key business functions and operations, in other words,these systems aim to improve the collaboration and information management (Liang & Huang, 2002).
A system to support collaboration and information management (ColIM) should be able to offer to theinstitution, similar to the concept of corporate portals defined in (Detlor, 2000), a shared information workspace; a communication space to negotiate collective interpretations and shared meanings; and a coordination
e front matter r 2006 Elsevier Ltd. All rights reserved.
nfomgt.2006.10.003
ing author. Tel.: +351 22 2094000; fax: +351 22 2094050.
ess: [email protected] (C. Sofia Pereira).
ARTICLE IN PRESSC. Sofia Pereira, Antonio Soares / International Journal of Information Management 27 (2007) 86–103 87
space to support cooperative work action, in other words, a shared information work space that facilitatesaccess to information content, organizational communications, and group collaboration.
Considering the model of collaboration 3C (communication, cooperation, coordination) (Ellis, Gibbs, &Rein, 1991; Borghoff & Schlichter, 2000), collaboration can be defined as ‘‘joint and coordinate action ofindividuals supported by communication’’. A collaborative tool should provide a space within whichparticipants can locate each other, exchange synchronous and asynchronous messages, share documents andapplications, share workflow, etc.
Much of the knowledge work in organizations require more and more flexible ways of collaboration(Agarwal, McParland, & Parry, 2002) in combination with powerful information management tools. Researchand Development (R&D) institutions, as prototypical of knowledge and project driven organizations, need, bytheir own nature, a high level of collaboration as much internally as with other institutions that share commongoals. Also, as knowledge intensive organizations, R&D institutions have information management as one ofthe fundamental business processes. The combination of these two dimensions, collaboration and informationmanagement assume, thus, an extremely important role in the success of the R&D institutions.
There are a number of key aspects that can affect the success of a collaborative venture, specifically in aresearch environment and these include: knowledge sharing (ColIM systems need to support the interactionamong the members of the several project teams, in way to share the knowledge and information of eachmember) and collaboration planning (ColIM systems should support actual work practices as the participantsconstruct them, rather than simply the officially documented practices) (Mulligan, O’Sullivan, & Beck, 2003).
The way that these principles are implemented depends of the particular context of each institution. Theperfect understanding of the requirements that support collaboration (coordination, communication andcooperation) and information management define the success of the ColIM system to be developed. The phaseof analysis in the information systems development is decisive.
To develop a system that guides the users in the sense of more effective collaborations, it is fundamental tovisualize and to understand the set of relationships that inside of the organizations facilitate or impede thecreation and transfer of information and the collaboration.
To have effective collaboration, it is necessary to consider with larger importance the informal networks.The collaboration, the share and the transfer of information and knowledge depend of the personal networkof each individual and of the individuals’ will to collaborate. So, the use of a method of organizational analysisthat allows to know the informal networks existents inside of the organization, in other words, that allows toknow the collaborative networks existent in the organization, will allow the analyst of the system tounderstand the processes of collaboration, information sharing and information flow inside of theorganization. After having diagnosed the problem, to propose a series of recommendations to improve thecollaboration and information management.
This paper, presents a new way for organizational analysis to improve the quality of the requirements ofColIM systems. The so-called SNetCol method, is based in the application of the social network analysisapproach, and it aims at deriving a richer set of requirements for collaborative information managementsystems. The rest of the paper is structured as follows:
Section 2, synthesize and discuss the characteristics of current IT platforms to implement ColIM systems.Two categories of systems are analysed: wiki systems and content management systems. Section 3, describes indetail the SNetCol method and the major concepts of the social network analysis to facilitate theunderstanding of the method. Section 4, presents the case of an R&D systems engineering institute where theSNetCol was applied and the results of the evaluation of technological options (wiki systems and contentmanagement systems), considering the collaborative requirements obtained using the SNetCol method.Section 5 provides some conclusions of this work and points out future research directions.
2. IT platforms for collaboration and information management systems
2.1. Modern platforms to support ColIM systems
Assuming that in R&D organizations there is a mix of highly structured environments as well as informaland spontaneous collaboration (Agarwal et al., 2002), the fundamental requirement for a collaborative
ARTICLE IN PRESSC. Sofia Pereira, Antonio Soares / International Journal of Information Management 27 (2007) 86–10388
information management system is that it supports efficiently both structured and unstructured collaboration/information. To specify a ColIM system for R&D institutions that, besides the fundamental requirement alsofacilitates social aspects of group formation, team collaboration and the exchange of information amonggroups two categories of systems were studied: content management systems and wiki systems.
The option for these systems as instruments of collaboration and information management in R&Dinstitutions was motivated by a set of characteristics that they present: (1) The flexibility presented by thesetype of systems that is one of the key points to be considered in the development of collaborative systems; (2)Easy adaptation to the user style (easy personalization), that it guarantees a good acceptance of the system bythese; (3) Capacity and execution facility of evolutionary modifications given the number of people dedicatedto the development of such systems; (4) Easy integration of new modules or functionalities in agreement withthe needs of each user; (5) The fact of offering predefined structures for the logic of the system that it reducesthe time of development; (6) The fact of supporting unstructured and structured collaboration; (7) Gather aset of collaborative functionalities and at the same time allow the efficient information and knowledgemanagement; (8) High number of plug-ins, add-ons and new products that are available frequently, that allowto increase or to improve whenever necessary, the functionalities offered; (9) Easy edition, actualization andshare of contents; (10) The ability to construct shared spaces to store documents, to exchange information andto work collaboratively in the execution of the various projects; (11) Allow the construction of an informationand knowledge base in the organization; (12) Supply to the project teams or groups of the organization an on-line shared space to store documents; to exchange information and to work collaboratively; (13) The simplicityto use these systems enables a more easier sharing of ideas; (14) Fast and simple actualization of the publishedinformation; (15) The diversified range of applications already implemented, in such different areas, usingthese two types of IT solutions.
2.2. Content management systems
A content management system (CMS) is a system that supports the creation, administration, distribution,publication and collection of information. CMSWatch define a CMS as a group of business rules and editorialprocesses applied to contents by people and organizations to align efforts of online publication with thebusiness goals (Byrne, 2004). The CMS is also known as ‘‘web content management systems’’ (WCMS), beingthese the more popularized, and being focused mainly in the on-line contents of the external pages andintranets in an organization. A CMS can be defined as a tool that offers a variety of techniques to create, toedit, to manage and to publish contents (in several formats), with a centralized set of rules, processes andworkflows that assure electronic contents coherent and validated. These systems collect the whole cycle of lifeof the pages of the sites (external and internal) of the organization, supplying simple tools for creation ofcontents, publication and archive, allows the administration of the site (internal and external) structure,published pages appearance and navigation supplied to the users (Robertson, 2003). Typically, it is used tofacilitate the information administration online. A CMS can also be used as a document management system,besides being an effective means for on-line collaboration and content creation. A characteristic particularlyinteresting of CMS is the fact that they allow publishing new and updated information frequently.
2.3. Wiki systems
Although not known as content management systems, wiki systems are another approach to publish on-lineinformation and a different way to collaborate. They can be characterized as ‘‘ultra-lightweight’’ contentmanagement systems (Mattison, 2003). Wikis are used for a diversified number of applications: some asdatabases for research and writing, as personal information manager, as collaborative tool between teams tocreate and maintain documents that need to be updated frequently, etc. Wikis can be used for almosteverything, because of their flexible structure.
Sauer et al. (2005), view wikis as a method for knowledge management and refer to a collaborativehypermedium that allows for continuous communication within a research team and the constant evolution ofcontent. Ward Cunningham and Bo Leuf, in the book ‘‘The Wiki Way’’, define wiki as ‘‘a freely expandablecollection of interlinked Web ‘pages,’ a hypertext system for storing and modifying information — a database,
ARTICLE IN PRESSC. Sofia Pereira, Antonio Soares / International Journal of Information Management 27 (2007) 86–103 89
where each page is easily editable by any user with a forms-capable Web browser client’’ (Mattison, 2003).(Neus & Scherf, 2005), define the wikis as web content management systems that allow collaborative creation,connection and edition of contents. The original concept of wiki does not impose any controls on who cancreate and edit pages. Like this, it allows the actualization and growth of contents continuously, and does notimpose any restriction encouraging multiple people to add content in a single page (Robertson, 2004). A wikiis a web site in which any one can put material and quickly do alterations without using complicatedcommands (Hof, 2004). This is the concept that created wikipedia, free on-line encyclopedia, where any onecan put information. However, actually these technologies allow controlling who can create, visualize and editpages, contrarily to the initial concept of wiki. This control becomes extremely important in certainapplications. Like this, a wiki is a server-based collaborative tool that allows the any authorized user to editpages and create new pages without learning any programming language (Chawner & Lewis, 2004).
Table 1 presents, in a synthesized way, some particular characteristics of each one of the approachedsolutions that allows to show some of the main existent differences among these systems, basing essentially onthe following works (Robertson, 2003; ARES, 2005; Byrne, 2004; Curran, Doherty, & Power, 2004; Wei et al.,2005; Bean & Hott, 2005; Lachkovies et al., 2004; Skiba, 2004; Robertson, 2004).
Although these two types of systems possess appropriate characteristics for systems ColIM, only after theanalysis phase and specification of requirements we will be in a condition to evaluate the applicability of thesetypes of systems in full detail as solutions for ColIM. The correct evaluation of the answer of each one of thesolutions as systems of ColIM depends directly on the results of the phase of analysis. Like this, the choice ofthe more appropriate method for organizational analysis is a key factor in the success of this process, becauseit influences largely the quality of the results obtained in the analysis of requirements and consequently theresults of the evaluation of different solutions as systems for ColIM. In the following section is presented themethod used for organizational analysis, SNetCol method.
3. Improving the requirements for collaborative information management by using SNA
3.1. Organizational analysis and the specification of collaboration requirements
The organizational analysis for specification of collaborative systems can be accomplished in several ways.A possible approach, referred by Davies et al. (2001), is the analysis of the information flow. Through the useof questionnaires and interviews, the authors analyse as the people receive, organize, create and structure theinformation in a personal and group perspectives aiming at discovering the different ways of coordination andexchange of information. According to these, the study also allowed to identify the causes and consequences offailures of the information flow that would focus the attention of the users’ requirements and in the functionalspecifications related with project of the system. With the data collected through the questionnaires andinterviews, information flow maps are drawn, in order to understand better the space and the range of theinteractions among the employees of the different departments (Davies et al., 2001). (Pinelle & Gutwin, 2000),explore the tendencies and techniques of groupware evaluation based on 45 papers presented in theconferences of ACM and CSCW, concluding that the existent methods do not answer to the needs ofgroupware evaluation. These authors propose a method to analyse and to specify group tasks. This methodallows the specification of scenarios of collaboration and tasks, considering the mechanisms of collaboration(explicit communication, implicit communication, coordination of actions, planning, monitorization,assistance and protection), union levels during the accomplishment of the tasks and variability in the tasksexecution, unlike the methods analysed that are focused in the mechanisms of collaboration, without worringabout the context in which such systems are used. DiMicco (2004) presents an approach for construction ofcollaboration tools, based on the observation and search of the existent problems inside the group relativelythe collaboration and then to build technologies that answer those problems. For evaluation of thecollaborative technology, efectivity uses methods existent in the field of the social psychology.
In this work, the methodology of SNA is used to make the organizational analysis for requirementsspecification of collaborative systems. The exploration of the social networks is described for (Hanneman,2001) as essential to find the collaborators capable to help in the problems resolution and to increase thecollaboration among individuals.
ARTICLE IN PRESS
Table 1
Some characteristics of the wiki systems and CMS for ColIM
Content management systems Wiki systems
Easy adaptation to the user style (easy personalization),
guaranteeing a good acceptance of the system;
In wiki systems all collaborators have writing and edition
permissions in the shared space (this is the philosophy of the wikis,
although actually can be implemented with access and edition
access, just as in the CMS). This increases the use facility and it
turns all members of the teams active/productive participants. This
facility or collaboration can turn this systems a powerful tool for
project management and collaborative writing.
A CMS supplies powerful methods for information search,
navigation and filtering capable to allow the users to find the
correct information in correct time, in other words, the information
that need and when need. The metadata can help in this process. A
CMS simplifies the capture and metadata manipulation.
The use of the wikis seems to encourage the free flow of
information between participants and the voluntary contribution
of all with additional knowledge/information. Although any users
can remove the work of other, this rarely happens.
They improve the precision and quality of information shared
inside of the organization: a CMS allows make available on-line
information updated, consistent and the high quality.
Wikis can be useful as shared social spaces among members of
teams that work remotely. As it is easy to edit a wikipage, it is
simple for the members of several teams create your homepage.
Theses homepages increase the social presence and they can be
considered as a valuable tool for communities construction.
Reduces the duplication of information: whenever possible the
information should be stored a single time and reused whenever
necessary. A CMS facilitates the content reuse.
Wikis have de advantage of allowing the exchange of ideas in small
project teams and to promote the discussion. Some authors make
the analogy of the wikis with a white screen, supplying potential for
a more creative environment and knowledge expansion in projects
management.
Supplying an interface easy to use for creation/edition of new
contents (WYSWIG interface), a CMS can remove many of the
barriers to disseminate information in the organization.
Although it supplies an interface easy to use for creation/edition of
contents, the wikis request that the users learn a syntax to
maximize the use of the formatting functionalities.
Allows decentralized contribution without losing the control of the
centralized process, through centralized workflow processes, rules
and approval processes.
The revolutionary idea of the wikis, one of the main advantages, is
that the users they can make alterations in the contents without
passing for an administrator bottleneck or revision board. The
quality control is done maintaining a historic of versions and
allowing old versions to be reestablished case the subsequent
version is not considered the best.
Through the workflow processes, rules and approval processes, a
CMS can to supply an effective audit trail that allows production
with responsibility and assures a content flow controlled around of
the internal processes.
In the wikis, the quality of the audit trail depends a lot on the
software wiki used.
A CMS improves the publication processes and supplies larger
transparency and responsibility in the contents publication on the
part of their producers or in agreement with the defined approval
flow, once ad hoc publication processes impede the effective
management and control.
Wikis supply an excellent way to annotate evolutions in the
projects.
C. Sofia Pereira, Antonio Soares / International Journal of Information Management 27 (2007) 86–10390
3.2. Social networks approaches in organizational analysis
In any process of information system development, the phase of requirements analysis and specification isfundamental for the success of the system in development meaning that, the right choice of the methods oforganizational analysis to use is a key factor in this process. Although some authors do not considerorganizational analysis as formally making part of the requirements engineering process (Sutcliffe, 2002), itcan be argued that at least for information systems strongly supporting collaborative processes theidentification and analysis of requirements is naturally intertwined with the analysis of several aspects such astasks interaction, work practices and business processes. To analyse the collaborative dimension of those
ARTICLE IN PRESSC. Sofia Pereira, Antonio Soares / International Journal of Information Management 27 (2007) 86–103 91
aspects, an approach that highlights this dimension is needed. As the analysis of the existing informalnetworks offers important insights that allow the analyst to present a set of recommendations (requirements),that can improve the collaboration, communication and information flow in the organization.
Social network analysis enables to visualize and understand the relationships that facilitate or hinder thecollaboration and sharing of information and knowledge in an organization. It is also possible to identify groups ofindividuals that carry out central roles, isolated groups or individuals, to detect information ‘‘bottlenecks’’, toidentify opportunities for improvement of the flow of information and knowledge, to improve the effectiveness ofthe formal communication channels, to increase the importance of the informal networks.
The process of social network analysis involves typically the use of questionnaires, interviews andobservation as instruments to collect data about the relationships between defined groups or networks(Barbedo, 2003; Garton, Haythornth, & Wellman, 1997; Groth, 2003; Palau, Montaner, & Lopez, 2004).Other methods of data collection can be used as, for instance, data collected through administrativeinformation, such as e-mail groups, group notification systems and organizational information. A vast rangeof measures of social network analysis exists. However, their use depends on the analysis context, being thatsome of the measures drive to similar results. In this study the following measures were selected: density,geodesic distance and diameter, maximum flow, degree centrality, closeness centrality, betweenness centralityand, for the subgroups identification, the use of the n-cliques concept. Through a careful study of the severalmeasures of social network analysis existent and the results that each one of them allows to obtain, themeasures presented, according to our perspective, are the most appropriate for the evaluation of the definedcollaboration criteria (see Table 7, annex 1).
3.3. Overview of the SNetCol method
As elaborated in Section 3.2, the SNA methodology allows to identify and understand the set ofrelationships that facilitate or hinder the transfer of information and knowledge in the organization in study.According (Anklam, 2005), that refers SNA as ‘‘organizational X-ray’’, this can be considered as a diagnosticand planning tool to increase the collaboration and consequently the information flow in the organization.Considering these reflections an approach to organizational analysis for collaborative informationmanagement requirements elicitation was developed. The main phases of the method are as follows:
(1)
Selection of a set of fundamental criteria to evaluate collaboration in an organizational context(collaboration criteria), departing from the model of collaboration 3C (mentioned in Section 1).(2)
Selection of a subset of SNA measures that are likely to indicate the degree of collaboration in the organization. (3) Establishment of a relationship set between the measures of SNA and the collaboration criteria. Thisrelationship was first established as a hypothesis, then validated in an action-research based cycle.
(4) Identification and specification of a set of general requirements for collaboration and for the contentmanagement system. These generic requirements are to be compiled from the professional and academicliterature on CMS implementation. Eventually some adaptation to the particular context of theorganization can be needed.
(5)
Establishment of a relationship set between the relationships established in (3) and the genericrequirements. Again, this was made first as a hypothesis formulation, then validated in an action-researchbased cycle.(6)
Data collection for the SNA analysis. (7) Calculation of the defined measures, using a SNA software. (8) Analysis of the calculated values evaluating the collaboration criteria in detail. In other words, toaccomplish the evaluation of the structural collaboration potential in the institution, through theinstantiation of the relationships between SNA measures and collaboration criteria.
(9)
Analysis of the results obtained in (8) using them in determining the details and importance of the generalcollaborative and CMS requirements.The first five steps form, the common basis for the application of the methodology. A third step may beincluded depending on the degree of customization required for the methodology, and it can be used with
ARTICLE IN PRESSC. Sofia Pereira, Antonio Soares / International Journal of Information Management 27 (2007) 86–10392
minimum customization ‘‘right from the box’’ if the collaborative and information management requirementsof the organization fall in a general case.
The ontology of the SNetCol method is represented in Fig. 1 and the workflow in Fig. 2.
3.4. Social network measures and collaboration criteria
Considering the 3C collaboration model referred in Section 1 and the results that can be collected throughSNA, the following table presents a set of collaboration criteria (CC), based essentially in Wellman (1997);Wurst, Hoegl, and Gemuenden (2001); Detlor (2000), that allows us to evaluate the conditions ofcollaboration in an organization and their connection with the measures of social network analysis treated inthe previous section. Table 2 presents some examples of the results of the application of phases 1–3 of theSNetCol method (see Fig. 2).
The following table shows some examples of the group of identified requirements for collaboration and forthe content management system and some examples of the relationship between the collaboration criteriapresented in Table 2 and the generic requirements. The results presented correspond to the application ofphases 4 and 5 of the SNetCol method (see Fig. 2).
The chosen requirements as example can support other collaboration criteria besides introducing them inthe previous section (Table 3).
Further on, in Table 5, these are classified in terms of importance degree. The importance degree allocatedto each requirement depends directly of the number of collaboration criteria that supports and the resultsobtained in the evaluation of each one of the criteria.
4. The case of an R&D systems engineering institute
4.1. The organizational setting
This study was accomplished in an R&D institution, an interface institution between the academic worldand the business world of the industry and the services, as well as the public administration, in the scope of the
organizational
analysis
collaborative information
management requirements
analysis
produces
inputs todeliver
are inputs to
are inputs tois a
deliverare used
to evaluateeneter in
determine in
detail are
implement implementcan beSNA
measures
tight
relationship
loose
relationshipCMS extended
functionlaities
CMS general
functionalities
CMS general
requirementscollaboration
requirements
relationship
between
SNA measures and
collaboration criteria
collaboration
criteriasocial network
analysis
collaborative information
management requirements
Fig. 1. Ontology of the SNetCol method.
ARTICLE IN PRESS
1. Selection the
collaboration criteria2. Selection the SNA
measures
3. Establishment of a
relationship between the
measures of SNA and
the collaboration criteria
4. Identification and
specification of a set of
general requirements for
collaboration and for the
content management
system
5. Establishment of
relationship set between
the relationships
estabilished in 3 and the
generic requirements
6. Data collection for the
SNA analysis
7. Calculation of the
difined SNA measures
8. Analysis of the
calculated values
evaluating the
collaboration criteria in
detail
9. Analysis of he results
obtained in 8 using them
in determining the
details and importance
of the general
collaborative and CMS
requirements
collaborationcriteria for R&D
Institutions
See table 2
Connectionbetween
collaborationcriteria andspecified
requirements
See table 3
Compilation of the
results of the
questionnaires
accomplished for
data collection, that
will be the input data
for the calculation of
the measures
throught the SNA
through the SNA
software.
See section 4.2
Results/values ofthe SNA measures
Evalutation of thecollaboration in the
organization in study,considering the results ofSNA and the relationship
with the deginedcollaboration criteria
collaboration
See table 4
Classification degeneral
requirements interms of importance(low, medium, high)
See table 5
Generalrequirements forcollaboration andfor CMS in R&D
institutions
See table 3
See table 2
Connection between thedefined collaboration
criteria and SNAmeasures selected
SNmeasures forcollaboration
criteriaevaluation
See table 8 (annex 3)
See table 7 (annex 1) and table 2
Fig. 2. Workflow of the SNetCol method.
C. Sofia Pereira, Antonio Soares / International Journal of Information Management 27 (2007) 86–103 93
information technologies, telecommunications and electronic, being devoted to scientific research activitiesand technological development, technology transfer, consultancy and advanced formation. This studyconsisted in the requirements identification for the implementation of a collaborative information systembased in the analysis of the current collaborative network of this institution. It considered the current meansfor collaboration, these being essentially the e-mail, phone and personal contact. The chosen working case wasone of the institute departments, whose competencies areas include topics in the domain of the operationsmanagement and information systems, applied to industrial companies and business cooperation networks. Atlevel of operations management it includes intra and inter-companies operations management, project andoptimization of production models, planning, production scheduling and control, logistics interns andperform monitorization and evaluation. The area of information systems includes requirements engineering,design and development of innovative systems, information systems planning and project management. Thisdepartment also has a vast experience in the area of the integration of industrial systems and consultancy,namely in industrial organization and business information systems, design project elaboration andmanagement, technological audits, benchmarking, technology transfer and formation.
4.2. Collaborative network analysis
Data collection was basically done through questionnaires and observation (corresponds to phase 6 of theSNetCol method, see Fig. 2). From the 50 collaborators making part of the department, 32 answered to thequestionnaire. The collected data was analysed using NETMINER 2.5.
The answers were validated by the authors by observation. This task was facilitated by the fact that theauthors belong to this department.
ARTICLE IN PRESS
Table 2
Relationships between collaboration criteria and SNA measures
Collaboration criteria SNA measuares Relation criterion—measure
Coordination Easy access and sharing of
information in a coordinated
way (through the attribution
of responsibilities about
information that is made
available) (CC5)
Density+degree
centrality+closeness
centrality+betweenness
centrality
The centrality measures allow identifying the position
in that the actors are in the network in relation to the
communication and exchanges of information. They
measure the number of walks of communication that
go by each node.) Actors with high values of centrality
are actors with potential easy access to the
information that circulates in the network. However,
the easy access and sharing of information should not
just depend on the individuals’ position in the
network, but also of the access and sharing rules
imposed by the institution in study, so that these rules
are accomplished is essential the existence high
coordination among individual contributions, even to
reduce the duplication of the shared information and
for us to maintain well structured in way to facilitate
your access. The easy access to the information
depends directly of the existent coordination among
the information shared already in the system and the
new information that intend to make available. For us
to guarantee the share in coordinated way of the
information inside of the institution it is important the
definition of rules and procedures that allow to
maintain the control on the flow of information in the
institution. The density and centrality measures allows
to evaluate the current conditions of access and
sharing of information and will be the starting point
for the definition of the rules of access and sharing of
information in the institution.
Communication Capacity to establish contact
inside teams and among
teams (CC1)
Density of the global
network+density of each
identified group
If the density of the network or sub-network (group) is
low, it means that the analyzed network is restrictive
when the actors have to establish relationships with
other actors. If the density is high, it means that the
relationship among actors can settle down easily.
Cooperation Support in the realization of
tasks in group (CC7)
Closeness
centrality+geodesic distance
The closeness centrality measures the number of steps
that an actor should undertake to enter in contact
with other in the network. The more central an actor
goes, closer it is of other actors and more quickly he
can enter in contact and to interact with other. It is a
measure of the autonomy and interdependence of the
actors, relatively to the control exercised by other.
Like this, actors with high closeness centrality have
larger facility to obtain support for accomplishment of
tasks, problems resolution and ideas generation. The
ideal will be a highly connected network, in other
words, with high density values and therefore values
of geodesic distances reduced among all pairs of actors
and values very similar of closeness for all actors. The
along with this criterion allows to identify the actors
that are in more advantageous positions to obtain
support in the accomplishment of tasks, in the
generation of ideas and problems resolution.
Support in the ideas
generation and problems
resolution at organizational
level and group (CC8)
C. Sofia Pereira, Antonio Soares / International Journal of Information Management 27 (2007) 86–10394
Members with relevant roles have been given a special care to answer to the questionnaire. The calculationof the several measures was made considering the directed connections and attributing weights to eachconnection. For a collaborator that responses that it interacts daily with another is considered a connectionwith weight 3, if weekly interaction weight 2, if monthly interaction weight 1 and if interaction does not exist,the weight is zero. So, a larger weight is attributed to the most frequent interactions.
ARTICLE IN PRESS
Table 3
Collaboration criteria versus generic requirements
Requirements Collaboration criteria
R2. The system should make possible the collaborators search. CC1 and CC6 (facility of location of potentials collaborators, for
problem resolution, accomplishment of a task or teams
constitution.).
R2.1. It should be possible the search of collaborators by name,
projects, interest areas, actuation areas, function, activities that
accomplishes in the organization, competences and accomplished
works.
R5. The system should make available a mechanism of powerful
and flexible of search.
CC2 (capacity of rapid dissemination of information at
organizational and teams levels.); CC3 (facility of information
organization and structuring at organizational level and group.),
CC4 (capacity to maintain updated all the collaborators about the
projects of the institution); CC5 and CC8.
R5.1. The system should allow the search of information for
metadata, project name, project type, title of the documents,
description, creation date, state, content type and author.
R5.2. The system should assure that whenever a content is created
or stored in the system is associated this a set of metadata.
R7. The system should assure/control the quality of the
information made available through the definition of approval
workflows and revision of the information to publish.
CC2; CC3 and CC5.
R7.1. By the user profile and of the permissions that this has, the
system automatically should authorize the publication or revision
of the information that this intends to publish.
R7.2. The system should generate an alert in the work area of the
user responsible for the approval/revision of contents, when these
are submitted for revision.
R10. The system should allow the easy publication of information
stored inside of the work areas of the groups, for other users.
CC2; CC4; CC5 and CC9 (high integration between project teams
and the project teams (this integration can be reached by the
communication of objectives and tasks of the teams, work plans,
accomplished adjustments, etc.))
R10.1. The manager of each project and/or the coordinator of the
unit will be the responsible for to define who has access to the
several types of contents published in the system.
R10.2. The access to the contents published should be based on
users profile (each profile will allow to define that each user can see
and to do).
C. Sofia Pereira, Antonio Soares / International Journal of Information Management 27 (2007) 86–103 95
4.2.1. Results of the social network analysis
The results of the analysis of the collaborative network of RDI-OU are presented in detail in Annex 3, theseresults correspond to the application of the phase 7 of the method (see Fig. 2). Considering the values obtainedfor the several measures of SNA used, main conclusions can be presented: (1) Reduced capacity to establishcontacts at global level (given that of the total of 2450 potential relationships, 262 effective connections exist,which leads to reduced values of density); (2) Are the old collaborators in the institution and that execute rolesof high importance, such as, responsible of area and responsible of project, the nodes that represent essentialroles in the communication capacity and transmission of resources in the network (this conclusion is the resultof the analysis of the cutpoint nodes and disconnection lines (bridges)); (3) Are the actors that execute roles ofhigh importance inside of the unit, SIAC and RUPA (unit coordinators), the actors with higher number ofalternative walks to establish contacts among the others actors in the network, these are actors recognized asmore influential and with larger flexibility of negotiation (this conclusion is based on the values obtained forthe maximum flow); (4) The values obtained for several centrality measures allow us to conclude that the mostcentral actors in the network in study are the unit coordinators, usually oscillating the largest values of thecentrality measures among these two actors, being these the actors with larger facility of access tothe information and communication, as well as the actors with larger autonomy, larger capacity to control theflow of information and communication in the network and the mainly responsible for the share of theinformation that circulates in the network. However, the obtained results show that neither of the most centralactors have close values of the absolute centrality (1), being until relatively moved away; (5) although in anyinstitution the existence of subgroups is always a reality, in R&D institutions, given your structure team based,
ARTICLE IN PRESS
Table 4
Results of the evaluation of the collaboration in RDI-OU
Collaborative criteria Results of SNA
Easy access and share of information in a coordinated way
(through the attribution of responsibilities about information that
is made available) (CC5)
The number of existent connections in the network, well as the
value of density obtained reveals difficulty of access and share of
information. Through the centrality measures it can concluding
that are unit coordinators, project managers and secretariat, the
actors with more easier access to the information, therefore that
play more important roles in the share of information. The actors
with more high values of in-degree centrality and betweenness are
that who have larger access to the information, the actors with
more high values of out-degree centrality and betweenness, are
actors with facility of information share. The actors with higher
values of closeness centrality are the actors that are closer of other
actors in the network; these assume an important role for to
facilitate the access and share of information in the RDI-OU.
However, none of the centrality measures reveals very high values;
the values of actors with higher results are close of 0,5.
Capacity to establish contact inside teams and among teams (CC1)The low values obtained for density of the network reveal reduced
communication potential among the collaborators of RDI-OU.
Support in the realization of tasks in group (CC7) The values of closeness centrality obtained show that the need of
mechanisms that facilitate the interaction among the several
collaborators for these to execute yours tasks of collaborative way.
It is verified strong connection among responsible of project and
the elements of the respective team. But it is verified weak
connection among the several collaborators. The density values
obtained for the several subgroups reveal strong connection among
the members that constitute the group, but need of mechanisms
that support the interaction among groups.
Support in the ideas generation and problems resolution at
organizational level and group (CC8)
C. Sofia Pereira, Antonio Soares / International Journal of Information Management 27 (2007) 86–10396
this becomes still more evident. Not forgetting that the structures team based represent the more recognizedway of collaboration, it is however, very important the information sharing and communication among theseveral teams to guarantee the success of the institution, assuming in this task the actors that participate inmore subgroups a fundamental role (the results of the analysis of n-clicks reveal that are the actors unitcoordinators, area coordinators and responsible of project, the actors that participate in more subgroups).
The clear predominance of the actors SIAC and RUPA in all accomplished analytical measures, denotestheir importance at organizational level, being their functions ‘‘unit coordinator’’, followed by NAU andSURA, actors whose main function is ‘‘responsible of projects.’’
4.2.2. Evaluation of the collaboration in RDI-OU
Considering the values obtained through the social network analysis and the existent relationship among theanalysis measures and the established collaboration criteria, it is presented in Table 4 an evaluationof the collaboration in RDI-OU, in other words, some results of phase 8 of SNetCol method are presented (seeFig. 2).
4.3. Collaboration requirements classification from the SNA results
In Table 5, are presented examples (an extract) of the identified requirements for the development of acollaborative system and its classification in terms of degree of importance (low, medium, high) consideringthe results obtained in the previous section and the organizational knowledge acquired through the experiencelived as members of the institution in study. This table shows some examples of the application of phase 9 ofSNetCol (see Fig. 2).
ARTICLE IN PRESS
Table 5
Requirements classification of a collaborative system in RDI-OU
Requirements Degree Justification
R2. The system should make possible the
collaborators search.
Medium The facility of establishing contact inside teams or among
teams and the localization of potentials collaborators for the
problems resolution, task accomplishment or to constitute
teams, are a collaboration criteria that can influence the
success and efficiency of the project teams inside the R&D
institutions. Considering the values obtained for the density
of the network, closeness centrality and geodesic distance it
is verified a reduced communication potential and proximity
among the several collaborators of RDI-OU. This
requirement can support the improvement of the referred
criteria.
R2.1. It should be possible the search of
collaborators by name, projects, interest areas,
actuation areas, function, activities that
accomplishes in the organization, competences
and accomplished works.
R5. The system should make available a
mechanism powerful and flexible of search.
High As can verify in Table 3, this requirement is essential to
support great part of the main collaboration criteria.
Besides, considering the analysis of the network described in
Table 4 that reveals little satisfactory results for any of these
collaboration criteria, the importance of this requirement
still increases more.
R5.1. The system should allow the search of
information for metadata, project name, project
type, title of the documents, description, creation
date, state, content type and author.
R5.2. The system should assure that whenever a
content is created or stored in the system is
associated this a set of metadata.
R7. The system should assure/control the quality
of the information made available through the
definition of approval workflows and revision of
the information to publish.
High For besides the easy share and spread of the information for
all inside of RDI-OU, it is equally important that guaranty
exists as for the quality of the information that is made
available. The facility of information organization and
structuring at organizational and group level it can influence
the quality of the same. Considering the values obtained for
the maximum flow, that reveal weak organization and
structuring of information in RDI-OU, this requirement
shows to be high importance in the case of RDI-OU.
R7.1. By the user profile and of the permissions
that this has, the system automatically should
authorize the publication or revision of the
information that this intends to publish.
R7.2. The system should generate an alert in the
work area of the user responsible for the approval/
revision of contents, when these are submitted for
revision.
C. Sofia Pereira, Antonio Soares / International Journal of Information Management 27 (2007) 86–103 97
5. Evaluation of IT systems options for the implementation of a collaborative information management system
A previous analysis of current IT solutions to implement collaborative information management leads toconsider content management systems and wiki systems as natural candidates. Both types of solutions can beused for different purposes inside of an organization, given the enlarged range of functionalities that present,as it is possible to confirm through the varied implementations of both solutions (Pereira, 2005).
From several comparison reports, and from the experience of this organizational unit, two open sourceapplications were chosen: Plone as a CMS system and TWiki as the wiki system. Table 6 synthetizes theevaluation of these systems against the requirements derived from the SColNet method.
The detailed evaluation of the answer of the content management system (Plone) and the wiki system(TWiki), to the requirements specified for the collaborative information system to develop, allows thefollowing conclusion: (1) TWiki can be characterized as a CMS given the characteristics that can beimplemented through this. (2) Wiki systems present some limitations in the answer to certain requirementsessentially due to the structuring lack and organization of the information. (3) Content management systems(Plone) possess workflow systems very powerful and flexible when compared with the wiki systems (Twiki). (4)Considering the search mechanisms, the wikis (TWiki) are much more limited than the CMS (Plone) (5) Thecontents edition using CMS can be very similar to the commercial word processors, allowing certain CMS theintegration with these systems, contrarily to the that happens with the wiki systems, where the edition has to
ARTICLE IN PRESS
Table 6
Analysis/Evaluation of CMS and Wikis
Requirements PLONE TWiki
R2. The system should make possible the collaborators
search.
| | (Search very limited when compared with Plone)
R2.1. It should be possible the search of collaborators by
name, projects, interest areas, actuation areas, function,
activities that accomplishes in the organization,
competences and accomplished works
R5. The system should make available a mechanism
powerful and flexible of search.
| (System of metadata very limited, that consists in the
construction of a table in the final of each wikipage)
R5.1. The system should allow the search of information
for metadata, project name, project type, title of the
documents, description, creation date, state, content
type and author.
R5.2. The system should assure that whenever a content
is created or stored in the system is associated this a set
of metadata.
R7. The system should assure/control the quality of the
information made available through the definition of
approval workflows and revision of the information to
publish.
| |/ (allows the definition of approval and revision
workflows, being the state of the document indicated in a
table in the end of the wikipage, and are defined as set of
users with permissions for approval and revision)
R7.1. By the user profile and of the permissions that this
has, the system automatically should authorize the
publication or revision of the information that this
intends to publish.
R7.2. The system should generate an alert in the work
area of the user responsible for the approval/revision of
contents, when these are submitted for revision.
R8. The system should supply mechanisms that allow
the easy edition and import of contents in the system, as
well as the publication of these in pre-defined areas.
| |/ (Answers partially to this requirement, allows the
templates construction based in tables, doesn’t allow
however that text is copied and paste directly of other
text editors, due to the formatting rules, very limited
when the structuring and organization of the contents
publication)
R8.1. The system should possess a text editor with
similar functionalities of the commercial word
processing software.
R8.2. The edition of any content should be based on
templates.
R8.3. The publication of any content should be mapped
for specific areas of the system considering the type of
content edited or imported.
R8.4. The system should allow text to be copied and glue
directly of other text editors, formatting it in agreement
with the defined styles.
C. Sofia Pereira, Antonio Soares / International Journal of Information Management 27 (2007) 86–10398
be made using a set of pre-defined formatting rules.(6) In CMS (Plone) the content edition is by default basedin templates, being these defined according to the content type that we intended to edit. In the wikis thetemplates use is possible, being however necessary the construction of these on the part of somebodyresponsible for the structuring of the system. In the case of TWiki, the templates construction is based ontables. (7) For presentation and publication of contents the CMS allow the definition of specific areas forpublication of contents considering the type of content to publish. In the case of the wikis all of the contentsare published in the wikipages (HTML pages) and the user is the editor of the content, responsible for theplace (wiki page) where the information will be presented. As the amount of information increases it becomesstill more difficult to maintain the information easily accessible for all. (8) Although TWiki answers the greatpart of the presented requirements, the functionalities for this presented as answer, are more restricted than
ARTICLE IN PRESSC. Sofia Pereira, Antonio Soares / International Journal of Information Management 27 (2007) 86–103 99
offered them by Plone. (9) The option by one or the other type of systems, is highly dependent on the numberand profile of the users of the system and of the amount and type of information that intend to manage. TheCMS offers an environment of collaboration that operates equally well so much for small as big teams. (10)The use of the wikis is more suitable for collaboration among a less number of users and for collaborativewriting.
6. Conclusions
The approach of organizational analysis used in this work for analysis and specification of requirements,SNetCol method, based essentially on the application of the methodology of social network analysis, isrevealed to have great applicability in the characterization of the collaboration, considering as starting point,criteria that define the ‘‘best practices’’ of collaboration. It makes possible to evaluate collaboration in anorganization, offering information on the informal organizational structure that coexists with the formal oneand that assumes an important role in the sharing of information and knowledge. It also allows also todetermine who effectively collaborates with whom inside of an organization by identifying the ‘‘weak points’’and the ‘‘strong points’’ of the collaboration inside of the institution and to classify in terms of importance therequirements of a collaborative information management system, leading to substantial improvements in theprocesses of managing and sharing information in a collaborative way.
Table 7
Social network analysis measures
Density The density of a social network is the quotient of the connections that exactly exists among the
network actors by the total of connections that could exist. It is an index of the communication
potential among the parts of the network and, therefore, of the amount and types of
information that can be exchanged theoretically (Palau, Montaner, & Lopez, 2004; Silva, 2003).
Geodesic distance and diameter The geodesic distance is the number of relations in the shortest possible walk from one actor to
another. The diameter of a network is the largest geodesic distance in the (connected) network
(Hanneman, 2001).
Maximum flow The maximum flow indicates the number of actors/nodes that allow to the source actor, by the
direct ties maintained for him with your neighborhood, access alternatives to the target actor in
the network.
Degree centrality Degree centrality is a measure which counts the number of ties an actor has. In case we are
dealing with a network where direction of ties is important, we can distinguish among in-degree
centrality and out-degree centrality. The in-degree centrality is the number of ties an actor
receives. The out-degree centrality is the number of ties which begin in the actor (Palau,
Montaner, & Lopez, 2004 Palau et al., 2004).
Closeness centrality In relation to the closeness, an actor is so more central as smaller the walk than he needs to
travel to reach the other actors of the network. Measure, in last analysis, her independence in
relation to the control of other. This measure proposed by Freeman (Hanneman, 2001), also
considers the indirect connections and sustains that the centrality is inversely proportional at the
geodesic distance of an actor in relation to all other actors of the network. For direct
connections it is necessary to do the distinction between in-closeness and out-closeness centrality
(Soares, 2002).
Betweenness centrality The interaction between two actors not adjacent in the network it can depend on other actors
that interpose among them. The betweenness centrality views an actor as being in a favored
position to the extent that the actor falls on the geodesic paths between other pairs of actors in
the network, if to play the role of intermediary in the interaction between those pair of actors
(Hanneman, 2001).
For subgroups identification:
n-Cliques
In a clique, all of the members have to have a direct connection with all other members of the
group. So, in way to make this concept including, useful and general, the n-click concept
appears, that it is defined as a maximum subgroup in which the largest geodesic distance among
two actors is not larger than ‘‘n’’.
ARTICLE IN PRESSC. Sofia Pereira, Antonio Soares / International Journal of Information Management 27 (2007) 86–103100
Relatively to the considered technological options, a CMS seems the most appropriate according to ouranalysis. However, both CMS and Wiki systems present high potential as solutions for collaboration andinformation management.
In terms of future work we have 3 main objectives: to apply the approach following in this work to differentenvironmental and organizational contexts in the extent of the specification of information systems forcollaboration and information and knowledge management; to develop a generic method of analysis andspecification of requirements based on this approach, defining broader criteria, appropriate the anyorganization type; and to apply the following approach in this work to the development of informationsystems for collaboration, information and knowledge management inter-organizational.
Appendix A
The social network analysis is show in Table 7.
Appendix B
Collaborative network RDI-OU (Fig. 3). The names associated to each nodes represented in the sociogramare fictitious, in way to guarantee the confidentiality of the data.
Fig. 3. Collaborative network RDI-OU.
ARTICLE IN PRESS
Table 8
Results of the analysis of the social network of RDI-OU
SNA measures Results of SNA
Density The calculated density of the network is low, indicating that only 10.8% of the potential of relationships
of the network are to be used. However, it is important to refer that, if the 18 collaborators that did not
answer to the questionnaire had answered, that would contribute to increase the density of the network.
Geodesic distance and
diameter
The mean geodesic distance is 2.208, it means that each actor needs 2 contacts on average to reach other
in the network. The diameter of the net is 4, in other words, he maximum distance that separates any two
actors in the net is 4. However, it can be verified through the map of the Fig. 1 the existence of isolated
actors.
Maximum flow The maximum flow of the network is 54, the medium maximum flow is 8.785, that means that each actor
has 9 alternatives on average to reach other actors in the network. This is a measure of the efficiency of
the network, therefore gives us the number of available walks for the exchange of resources.
Degree centrality Relatively to the in-degree centrality, the most central actor is SIAC (0.408), following by SARP (0.367)
and SURA and RUPA (0.286). As for the out-degree centrality we have MAR (0.592), following by
SOIA and RUPA (0.327). This measure presents in-degree index of centralization 30,61% and out-
degree index of centralization 49.35%. This centralization index is a comparative parameter that it
allows to evaluate the degree of total centralization of this graf, relatively to a referencial value based on
the topology in star, which it is the graf model in that the centralization is perfect.
Closeness centrality For the in-closeness centrality we have the actor SIAC (0.444), following by SARP (0.409), SURA and
RUPA (0.381) and NAU (0.348). For the out-closeness centrality we have RUPA (0.531), SIAC (0.517),
following by SOIA (0.51) and SURA (0.484). The index of the in-closeness centralization is 41.714%,
and the index of out-closeness centralization is 82.595%, that it constitutes high values of global
centrality. The obtained mean values are 0.242 (in-closeness centrality) and 0.239 (out-closeness
centrality).
Betweenness centrality The actors with larger values of betweenness centrality are SIAC (0.096), following by RUPA (0.095),
OTIR (0.079) and NAU (0.071). The centralization index of betweenness is 8.515%.
For subgroups identification:
n-Cliques
They were found 15 particularly united subgroups. It is verified that several actors belong to more than a
group simultaneously. For instance, actor SIAC integrates all identified subgroups, actors RUPA and
SURA do part of 14 of the identified subgroups, following by: NAU (12); OLOS and SGAN (11); PAS
and DAM (9); RADE and OTIR (8); SESI and NOIA (6); RAU and ARES (5); LOLA, DAIA and
MAR (4); SOIA, ALOC and ROLA (2) and finally AAAS, ALU, NOA, OAIP, UXA, GARI, PEL (1).
C. Sofia Pereira, Antonio Soares / International Journal of Information Management 27 (2007) 86–103 101
Appendix C
Table 8 presents the results of the analysis of the collaborative network of RDI-OU. Relatively to thecentrality measures, it would be interesting to present the values of centrality of each node. However, as thenetwork has a considerable dimension, only the actors with larger centrality values and larger centralizationindexes in each centrality measures will be considered. It is important to remember that these values oscillateamong 0 (periphery) and 1 (center). For the identification of the existent subgroups in the network the n-clickconcept was used. The chosen value for n was ‘‘2’’, in other words, were returned all actors cohesively united inthe network for a distance equal or less than two geodesic walks. It is important to refer that for theidentification of the subgroups were only considered the confirmed interactions by both the two actorsinvolved. With the obtained results we tried to determine whose actors participate in more than a clique,evaluating the intervention degree and those actors and the weight in the institution, as well as theirimportance in the integration of the various teams.
References
Agarwal, D., McParland, C., & Parry, M. (2002). Supporting collaborative computing and interaction. In Proceedings of the grace hopper
celebration of women in computing 2002 conference. Vancouver, Canada, October 9–12.
ARTICLE IN PRESSC. Sofia Pereira, Antonio Soares / International Journal of Information Management 27 (2007) 86–103102
Anklam, P. (2005). Social network analysis. Avaliable at /http://www.byeday.net/sna/index.htmlS.
Ares. (2005). Fact sheet—content management systems. Sydney: ARES Interactive 2005.
Bean, L., & Hott, D. (2005). Wiki: A speedy new tool to manage projects. Journal of Corporate Accounting and Finance, 16(5),
3–8.
Borghoff, U.M., & Schlichter, J.H. (2000). Computer-supported cooperative work: Introduction to distributed applications. New York, USA:
Springer.
Byrne, T. (2004). Web content management products & practices. CMS watch, version 6.0. The CMS report (enterprise edition).
Chawner, B. & Lewis, P. (2004). WikiWikiWebs: New ways of interacting in a web environment. Handout from LITA forum. St. Louis,
Missouri, October 7–10.
Curran, K., Doherty, K., & Power, R. (2004). WikiWikiWeb as a tool for collaboration. Information Technology Journal, 3(2),
206–210.
Davies, J. G., et al. (2001). Creating shared information spaces to support collaborative design work. Information Systems Frontiers, 3(3),
377–392.
Detlor, B. (2000). The corporate portal as information infrastructure: Towards a framework for portal design. International Journal of
Information Management, 20, 91–101.
Dimicco, J.M. (2004). Methodologies for evaluating collaboration in collocated environments. CSCW workshop on co-located
collaboration.
Ellis, C., Gibbs, S., & Rein, G. (1991). Groupware: Some issues and experiences. Communications of the ACM, 34(1), 39–58.
Garton, L., Haythornth, C., & Wellman, B. (1997). Studying online social networks. Journal of Computer Mediated Communication, 3, 1.
Groth, K. (2003). Using social networks for knowledge management. Departament of Numerical Analysis and Computing Science, Royal
Institue of Technology, Sweden. In Eighth European conference of computer-supported cooperative work, ECSCW’03.
Hanneman, R. (2001). Introduction to Social Network Methods. Department of Sociology, University of California: Riverside.
Hof, R. (2004). Something Wiki this way comes, They’re Web sites anyone can edit: And they could transform corporate America.
BusinessWeek, June 7. Avaliable at /http://www.businessweek.com/magazine/content/04_23/b3886138.htmS.
Lachkovies, A., et al. (2004). Innovative use of internet-based collaboration tools and methods in an institutional context. Third PAN-
commonwealth forum on open learning (PCF3). New Zealand: Dunedin.
Liang, W., & Huang, C. (2002). The agent-based collaboration information system of product development. International Journal of
Information Management, 22, 211–224.
Mattison, D. (2003). Quickiwiki, Swiki, Twiki, Zwiki and the Plone Wars Wiki as a PIM and collaborative content tool. Searcher, 11:4
(April). Avaliable at /http://www.infotoday.com/searcher/apr03/mattison.shtmlS.
Mulligan, D., O’Sullivan, D, & Beck, R (2003). Collaborative information system for university based research institutes. computer integrated
manufacturing research unit (CIMRU). Galway: National University of Ireland.
Neus, A., & Scherf, P. (2005). Opening minds: Cultural change with the introduction of open-source collaboration methods. IBM Systems
Journal, 44(2), 215.
Palau, J., Montaner, M., & Lopez, B. (2004). Collaboration analysis in recommender systems using social networks. Institut d’Informatica i
Aplicacions, Agents Research Laboratory. Espanha: Universitat de Girona.
Pereira, C. (2005). Gestao de informac- ao numa instituic- ao de I&D: Uma abordagem por redes colaborativas. Ph.M. Thesis, Faculdade de
Engenharia da Universidade do Porto, Porto.
Pinelle D., & Gutwin C. (2000). A review of groupware evaluations. In Proceedings of the nineth IEEE international workshops on enabling
technologies: Infrastructure for collaborative enterprises.
Robertson, J. (2003) So, what is a content management system? KM Column, June 3. Avaliable at /http://www.steptwo.com.au/papers/
kmc_what/S.
Robertson, J. (2004). intranets and knowledge sharing. KM Column, May 5. Avaliable at /http://www.steptwo.com.au/papers/
kmc_intranetsknowledge/S.
Sauer, M., et al. (2005). ‘‘Blogs’’ and ‘‘Wikis’’ are valuable software tools for communication within research groups. Artificial Organs,
29(1), 82–83.
Silva, M. (2003). Redes sociais intraorganizacionais informais e gestao: Um estudo nas areas de manutenc- ao e operac- ao da planta HICO-8,
CAMAC- ARI, BA. Ph.M. Thesis, Escola de Administrac- ao, Universidade Federal da Bahia.
Skiba, D. (2004). Emerging technologies center. Nursing Education Perspectives, 26(5), 310–311.
Soares, W. (2002). da metafora a substancia: redes sociais, redes migratorias e migrac- ao nacional e internacional em valadares e ipatinga.
Ph.D. Thesis, Faculdade de Ciencias Economicas da Universidade Federal de Minas Gerais.
Sutcliffe, A. (2002). User-centred requirements engineering: Theory and Practice. UK: Springer-Verlag, ISBN: 1-85233-517-3.
Wei, C., et al. (2005). Wikis for supporting distributed collaborative writing. In: Proceedings of the society for technical communication
52nd anual conference, Seattle, USA.
Wellman, B. (1997). What is social network analysis? Toronto.
Wurst, K., Hoegl, M., & Gemuenden, H. (2001). Collaboration Within and Between Teams in Multi-Team R&D Projects. Management of
engineering and technology. In PICMET ‘01, Portland international conference on volume supplement, 2, 545–552.
Carla Sofia Pereira is a doctoral student at the Faculty of Engineering of the University of Porto, Portugal. She is currently a lecturer in
the School of Management and Technology of Felgueiras of the Polytechnic Institute of Porto and a researcher in the Institute for Systems
and Computer Engineering of Porto. She received her M.Sc. degree in Electrical and Computers Engineering from the Faculty of
ARTICLE IN PRESSC. Sofia Pereira, Antonio Soares / International Journal of Information Management 27 (2007) 86–103 103
Engineering of the University of Porto. Her research interests are analysis and specification of information systems, collaborative
information systems, information management, knowledge management and Enterprise Cooperation Networks.
Antonio Lucas Soares is an Associate Professor of Informatics at the Faculty of Engineering of the University of Porto. He is also a Senior
Researcher at INESC Porto were he coordinates a research group on Information and Knowledge Management in Collaborative
Networks. Its current research interests are ‘‘social aspects of semantic resources development and use’’; ‘‘community informatics’’;
‘‘multi-enterprise ontologies’’; ‘‘collaborative working environments’’.
Top Related