CSCW systems in virtual environments:a general development framework

W. J. Sarmiento∗‡ and Cesar A. Collazos†

∗IDIS Research Group, Universidad del Cauca, wilsonj@unicauca.edu.co†Full Professor, IDIS Research Group, Universidad del Cauca, ccollazo@unicauca.edu.co

‡Multimedia Research Group, Universidad Militar Nueva Granada, wilson.sarmiento@unimilitar.edu.co

Abstract—Collaborative Virtual Environments (CVEs), de-fined as systems that allow performing some collaborative tasksin a virtual world, were proposed since 1990. From the begin-ning, these systems were developed as particular applications ofComputer-Supported Cooperative Work (CSCW) and virtualreality concepts about problems with specific restrictions onboth collaborative work and technological (hardware andsoftware) requirements. Therefore, design guidelines whichcould be applied to new developments have not been proposed.This paper presents a general framework for designing anddeveloping CSCW systems in virtual 3D environments, whichintegrates new methodologies that provide clear and formaltechniques to develop CSCW, interactive and 3D systems

Keywords-Virtual environments; groupware; framework anddesign methodology.

I. INTRODUCTION

Technological development aims to supply computationaltools for supporting tasks developed by humans. Clearexamples of this fact are the Computer-Aided Design (CAD)systems, which provide tools to facilitate the work of de-signer, whilst they perform the creative labor. Similarly,Computer-Supported Cooperative Work (CSCW) systemsprovide computational tools that looking for supporting andfacilitating the work group, i.e. tasks that are carried outcooperatively by various users.

Therefore, CSCW requires more than computational tools,because it requires understanding all aspects of the workgroup. Consequently, several areas such as psychology, soci-ology, organizational theory, among others are also involvedin the CSCW research [1]. Research on these areas allowidentify some key processes in the work group, which mustbe taking into account for developing CSCW systems: 1)Communication, between team members; 2) coordination,under a set of clear rules, and 3) information sharing in ademocratic mode [1], [2]. Other important components inCSCW are the computational tools; they are the channelsthat allow the users working as a team for achieving acommon goal. For doing so, special characteristics should beincluded in those computational tools, which are commonlycalled groupware o simply CSCW systems [1]. Therefore,the development of CSCW systems require to take intoaccount the key processes of the work group, therefore,

firstable, it is necessary to define particular features of thisprocess into a specific task or work [1].

It is known that observation is one of the most useful toolsfor characterizing any process done by a user, but, not anyobservation process could allow obtaining valid conclusions,formal and documented process is also required. In theparticular case of working group, observing the group ofpeople working and cooperating in a natural environmentallows to identify mechanisms that facilitate to achieve theobjectives of the cooperative tasks. To underline a text, towrite footnotes, to draft some ideas, are some examples ofthose mechanisms. IIn CSCW systems these mechanisms areknown as ‘sharing artifacts”, and could be implemented intwo ways either attempting to reproduce a real context orproposing new mechanisms according to the used technol-ogy.

The sharing artifacts are used to facilitate the key pro-cesses of work group, and also used for achieving the maingoal of the CSCW systems, as generating awareness in acooperative task. The users must have awareness that arepart of a team, consequently that their actions affect to others[1]. Also users must have awareness about how technologicalsystem mediates between them, by which it is possible toobserve latency in the action of the other users, associatedto communication channel restrictions [2]. Finally, the usersmust have awareness that they are executing a task and needinformation about the task activity [1].

However, technological restrictions are in some casesoppose to the cooperative work of users; for example, in anatural environment, communication process involve naturalinteraction between team members such as tacit and no ver-bal messages; coordination process includes monitoring therelevance of the participant feels (enthusiasm, compromise,etc.) and information sharing does not have time or physicspace limitations, which must be done in the dinner or inthe corridor.

The Collaborative Virtual Environments (CVE) has beenproposed as an alternative for improving those aspects inCSCW systems. A CVE is a system that allows carryingout a cooperative task in a 3D virtual world, which providesthe possibility of involve several aspects of a natural col-

laborative scenario in a CSCW system. These systemswereproposed since 1990, and its advantages have been exposedand verified by several applications [3]. A simple but revo-lutionary aspect of the CVE is the user interaction, which iscarried out by a graphic embodiment, called avatar. The useof avatars provides alternatives of visual appearance, facialexpressions and physic interactions near to reality that allowsinvolving gestural and no tacit communication options [4].Other relevant aspect is the possibility to offer a spacepotentially infinite, a virtual world, where the collaborativework can be carried out. This virtual world can be designedfor generating empathy with the working team and includeseveral spaces for the task may be done [3].

Likewise, the sharing artifacts can be designed followingmetaphors near to reality; for example, it is possible seea team member doing a presentation in an environmentsimilar to an office, or see how the avatar of a teammember disappears when he/she is disconnected, amongothers [5]. On the other hand, many other mechanismshave been developed for improving the sensation of realismsuch as enveloping visualization devices, which increasesthe immersive sensation improving the mode as users areinvolved with the task [3]; visual tools for monitoring thework of the group and following sharing tasks, personalizeddetail views [4] and, more recently, haptic and multimodalinterfaces [5].

As it expected, the introduction of those new aspectsincreases the complexity of those systems, resulting inmore and more design requirements that include graphicdesign, user interfaces, objects and 3D navigation spaceschoice, usability and ergonomic conditions, and many otherfeatures of the collaborative work. Previous works haveproposed software architectures for building of CVE [6],[5], [7], aiming its proposals to software and hardwarecomponents and intercommunication between these, leavingthe side important issues as design of collaborative work, 3Dinteraction, sharing artifacts, among others. Other proposalshave been centered in 3D interaction models in VirtualEnvironments (VE) [8], [9];emphasized in the problems of3D navigation, selection, manipulation and system control,and in the concepts of 3D awareness. Also, there are worksthat present the CVE systems, its architecture, advantagesand features [10], [11], [12], ], even some describe thedesign process used [13], but this experience are not fullyreplicable, as a guide that can be applied as a formal method.Therefore, it is necessary defining a methodology that guidethe process of development of a CVE, from the design, to theevaluation, passing through to the development of softwareas such. This methodology should be the proposals integra-tion around the design and development so much of systemCSCW as applications 3D, basing in CVE concepts, thispaper proposes a general framework as an approximation ofthis methodology.

The framework proposed in this paper allow the design

of CSCW systems in virtual environments, which attemptto integrate new methodologies that provide clear and for-mal techniques to develop CSCW and interactive and 3Dsystems. The next section includes a brief review of relatedworks. In the section 3 details of the proposed framework aredescribed, the next section shows framework implementationdetails through an example, and the last section presentsgeneral conclusions and some reflections about future work.

II. RELATED WORKS

This section shows the background about design anddevelopment of CVE, the first part presents a descriptionof more representative CVE, and the next explains otherproposals in order to design these systems.

A. CVE developed examples

Since the CVE were proposed, some examples haveshowed their design and development. Classic examples asMASSIVE [11], Spline [12] y DIVE [14], were the fiststhat include the main concepts as multiuser support, avatars,interactions with world objects, etc. But the more importantcontribution of these classic systems was the basics conceptsof 3D awareness: aura, focus and nimbus. Aura is defined asa volume around an object that border its presence in virtualworld and allows the interaction with others [10], [14], focusis a region were an observer can put its attention, and nimbusis a region surrounding an object that limit its observability[11], [15]. Thus, interaction between two objects could bepossible if the auras are overlapping, and so an observercan put his/her attention in an object. Previous examplesshowed the advantages and the potential of the CVE, butdo not allow to transfer the experience in other contextsbecause no describe the process used for their development;furthermore, these systems have been designed for generalpurposes with technological exploration as main target.

Some more recent examples have extended the function-ality of systems developed previously; as sTeam3D, whichis a web X3D client that allows transparent connection toa sTeam server. This system takes advantage of sTeam, anopen source groupware that provides some mechanisms tocommunication and cooperation, so much synchronous asasynchronous mode, and combines these mechanisms withhypermedia document management [16]. A similar exampleis SLMeeting, which implements some general purposecollaborative activities in Second Life platform, as agenda,meeting chat board, booking list, among others. This systemis supported chiefly by ability already included in SecondLife for creation, administration, navigation and interactionwith virtual worlds [17]. However, these systems not presenta formal guide for integrating CSCW systems with VEapplications.

MaDViWorld is other representative example; this systemis a framework for the development of VE with some col-laborative features [18]. This work allows manage multiple

users session in a virtual room through avatars, where theusers can be interacting with objects inside room, evenallows multiple virtual room simultaneously. However, thismodel is only a viewpoint of the user sessions management,proposing the use of “virtual world paradigm” against posi-tion to the traditional “document paradigm”. MaDViWorlddoes not offer any support to management or design of theVE, avatars and 3D interaction, since the authors indicateliterally “These aspects, however, are not our first preoccu-pation and are rather considered as improvements that couldbe added later within the system implementation layer of theframework” [18].

B. Other methodological approaches

Benford et al. presented the design of a CVE; they usedthe ethnography as methodology for the formal observationprocess of users. The report published describes the resultsof ethnographic study carried to the interior of TRC Com-pany, where they identified that the documents generated bythe company were the core of the collaborative processes.Thus, documents were modeled as “shared objects” overwhich they designed the shared artifacts. The authors showthe importance of the design of the collaborative processesas central point of CVE development [13]. Although thereport presents some recommendations of design, this not isa methodology of design and development.

Some work proposes general architectures in order todevelop CVE. Dai et al suggest a model that integratesthe conference function of NetMeeting with VE capabili-ties using VRML [6], the proposal describe details of thesoftware components, which are using COM as interface forthe system integration. the proposal describe details of thesoftware components, which are using COM as interfacefor the system integration. Gomes et al. propose an inte-gration framework for extended CVE with third party tool,the proposal core is “integration module” between CoLab,Platine and VNet, where CoLab allows collaborative capa-bilities, Platine provides communication tools (chat, audioand videoconference) and VNet is a free multi-user VRMLsystem [5]. These proposals show systems architecture butdo not include guidelines to design or to develop CVE.

Other proposals aim to design 3D interaction in CVE, forexample, Otmane et al. present a formal method to specifyand mathematical modeling three 3D interaction functions:navigation, selection and manipulation, and their relationwith the specific concepts of 3D awareness (aura, nimbusand focus) [8]. Hrimech and Merienne suggest a method forevaluating the impact of 3D interaction in CVE, in order todevelop a guide of design. They proposed three factors for asystem evaluation: co-presence, involvement/awareness andcollaborative effort apply to the 3D interaction functions.The factors are evaluated with a set of metrics acquired byuser observation, monitoring system and applied question-naire [9].

Nonetheless, there are methodologies, successful tested,directed to design and development CSCW systems; asSHASTRA [19], cooperation scenarios [20], designing withethnography [21], AMENITES [22], ontology-driven anal-ysis [23], among others. TThese methodologies, each onefollowing itself approach, show a formal process in order tobuilding groupware, with common emphasis: collaborationwork as the main goal to CSCW; but these methodologieshave not been proposed to develop CVE systems.

In other hand, the development of some 3D software needto consider algorithms and techniques of real time render,3D interaction, and specific data representation, in the soft-ware engineering process [24]. Also, in some particular case,i.e. game 3D development, activities as conceptual design,write history, modeling, animation of characters and world,must be integrating in the formal development process [25].

Some methods de facto have been created in the game de-velopment industry; there are modifications of agile method-ologies where conceptual and artistic activities have beenincluded in design process [24], [25]. In order to integratethese activities in the software engineering process, someapproach proposes to use the Unified Modeling Language(UML) for documenting 3D modeling and animation ac-tivities [26]; and other proposals use no rigid documentsas character sheets, design document, concept document,project plan, etc. [24], [25].

III. PROPOSED FRAMEWORK

CollaborationModel

Softwaredevelopment

Usersobservation

Software

Requirements

Colla

bora

tion

Mon

itorin

gAwar

enes

s

Mon

itorin

g 2

1

3

4

5

6

7

Visualization and Interaction Model

3DDesign

Figure 1. Sketch of proposed framework.

Taking into account the specific characteristics of CSCWsystems in VE explained above, our framework proposes toinvolve aspects related to the design of both 3D software

and CSCW systems in a common approach. Therefore,our framework preserves the main goal of CSCW, namedcollaborative work, and the main features of 3D softwarei.e. the well-known interaction possibilities. The figure 1illustrates a sketch of the proposed frame work, which is acycle composed of seven stages: requirements engineering,3D design, collaborative model, interaction and visualizationmodel, software development, collaboration monitoring andawareness monitoring, which should be performed itera-tively. Note that, the proposed framework starts and finalizeswith the “User observation” process, which must be the coreof any development process in systems that involve userinteraction.

A. Software requirements (Step À)

As it is usual, requirements engineering provides meth-ods and techniques that allows a formal and standardizedelicitation, specification and analysis of needs of users [22],[23]. Herein, we emphasize that this process must be basedon the user observation; particularly CSCW systems requireobservation of a team collaborating for developing a specifictask in the “real conditions”.

B. 3D design (Step Á)

The main advantage of work in a virtual world isthe amount of communication and interaction possibilities.However, these possibilities only can be provided if VEdesigns generate empathy with the users. For example,development of video games includes a conceptual design(scenarios, characters and history) that allows reaching ahard sense of immersion in the gamers [25]. In order to applythis concept, the proposed framework defines a specific stepfor obtaining a high level, or conceptual, 3D design. Thisconceptual design must to include scenarios, avatars, sharedartifacts and interaction modes and to be results of a formaluser observation process. Furthermore, these designs must beintegrated with the software development process by usingdocumentation and formal languages such as UML [26].

C. Collaboration Model (Step Â)

When user requirements and 3D interaction modes havebeen specified, the collaboration model defines a formalschema that abstracts the key processes of work group(communication, coordination and sharing) in the domainproblem. Thus, this step delivers a dynamic collaborationmodel that includes workflow, user roles, assigning-splittingtask and negotiation processes. Additionally, this model alsomust to define low level processes such as session users, in-formation persistence, concurrence and synchronism, amongothers. Taking into account that the collaboration modelis the main aspect of most of the CSCW developmentmethodologies, this stage could include approaches proposedby any of those methodologies [19], [22], [21].

D. Visualization and Interaction Model (Step Ã)

3D visualization and interaction model should be designedbased on the dynamic collaboration model defined in theprevious stage. Whilst the 3D design was focused on theconceptual design of the CVE, the visualization and interac-tion model aims to specify how the users will interact in thatVE. For doing so, it is essential to define the mechanisms forgenerating 3D awareness such as aurea, nimbus and focus,which define the characteristics of relevant VE elements thatshould be designed [3]. First, it is necessary to define thenavigation and object selection schemes, for instance avatarmovement options (modes and roads), task visualizationlevels, and view customization alternatives, among others.The tacit and explicit avatar communication modes, thespecific sharing artifacts and how they are used must bealso specified and designed in this stage.

This stage includes also activities related to 3D modelingand animation of VE elements, which must be coherent withthe conceptual 3D design defined in the step Á. Additionally,hardware devices that allow that users visualize and interactwith the 3D virtual world must be choices. Although, deviceselection attempt to achieve the interaction requirements, it ispossible that new software requirements should be specifiedfor supporting that in the system

E. Software development (Step Ä)

Once the particular application design models (3D, col-laboration, visualization and interaction models) have beenelaborated, computational applications should be imple-mented. For doing so, several methodologies have beenproposed, which generally compress both architectural de-sign and codification. In the architectural design, low levelrepresentation of VE elements (3D objects, persistent data,collaborative tasks, etc.) and their dynamic behavior shouldbe modeled [24]. Then, this architecture is used a base forbuilding a software application that should to achieve all re-quirements defined in previous stages. Herein, it is importantto emphasize that the whole framework is proposed as aniterative process. So, software development will also be aniterative process, when each iteration allows fulfilling a setof requirements, resulting in several software prototypes.

F. Collaboration Monitoring (Step Å)

One of the most important steps in software developmentis the test process, which allows introducing user feedbackfor improving details in the next iteration. As was describedin the previous sections, requirements of CSCW systemsthere are not only related to functional and interface capabil-ities but it is also relevant to assess the accomplishment of itsmain goal i.e. to facilitate the development of collaborativetasks. For achieving that, the framework specifics an activitynamed collaboration monitoring, which should be carried outmean of a formal and documented user observation processwhen they are interacting with the system. Measurement

of collaboration degree is yet an open research problem.Some aspects that should be evaluated in the CSCW systemsare the cohesion between team members, the kind andquantity of messages and information shared between them,the efficiency and effectiveness task developments, amongothers [19].

G. Awareness Monitoring (Step Æ)

Another fundamental aspect that must be evaluated is thedegree of user awareness. This monitoring process finds tomeasuring the relevance of the interaction and visualizationschemes provided by the VE for generating a high level ofuser awareness in different levels such as self awareness ina collaborative team, awareness of the other and the workgroup. In this stage some questions should be asked, forinstance: how the interaction and visualization devices arerelevant to these awareness senses? what is the comfort ofusers with the avatars and world aspects?, what is the degreeof user fatigue produced by the interaction and visualizationdevices?, what is the degree of ergonomics and usabilityof the interaction models?, etc. It is to say; in this stagethe quality of experience in the collaborative work shouldbe determined [8], [9]. Newly, development of objectivemeasures that allow evaluating the user awareness is alsoan open research problem.

IV. IMPLEMENTATION DETAILS

This section illustrates the main implementation details ofthe proposed framework using a simple example. We assumethat it is necessary to develop a CVE for promoting theworking group in a company in order to improve the laborclimate. For doing so, the CVE must to achieve the followingcharacteristics 1.

It is necessary to design a contest in a virtual world.Several teams will carry out the game, each of themcomposed by four gamers. The main goal will be to foundten keys that will be hidden in ten virtual rooms, whichwill be unconnected. Each virtual room will has severalboxes, one of them containing the searched key. Ten clues,one by virtual room, will be provided in order to facilitatethe searching task. The team that founds the ten keys inthe smaller time will win the game. Game start by placingthree team members in three different virtual rooms, thefourth member will act as an observer that will be ableto interact with the others using a communication artifactsuch as a chat room. Gamers will have two minutes forexploring the room, after that time they will be translatedto other rooms and roles will be changed i.e. other gamerwill be selected as observer. Therefore, the observer rolewill be essential for supporting the exploration task inrooms observed or visited previously by him

Text 1: General CVE description.

The proposed activity satisfies three characteristics re-quired by a collaborative task: 1) egalitarian participation,2) individual responsibility and 3) positive interdependence.The last one is understood as individual success given bythe group success [27]. The proposed framework suggestto carried out an ethnographical process i.e. to observethe users in real environments, for identifying the softwarerequirements [20], [22], [23] (step À ). However, to make ourexplanation easier, in this case software requirements wereexplicitly defined. Based on the CVE description providedin the text 1, a requirement analysis process is carried out.As any software development process, functional and non-functional requirements should be identified and specifiedi.e. formalized using a specification template and use casediagrams, figure 2 shows the user case diagram of exampleapplication.

Explorer

Observer

Find box

Guideexporer

Walk inroom

Open box

Sendmessage <<include>>

<<include>>

Figure 2. User case diagram

In order to improve the system acceptance by the users,it is necessary to guarantee a visual coherence that allowsgenerating a good empathy user-system. For doing so, theproposed framework suggest to develop the 3D design stage(step Á ) which involve the conceptual design of 3D com-ponents that will define the avatar visual appearance. Figure3 shows an initial conceptual design in character sheets.In this case a simple but friendly appearance is defined.It is expected that proposed design generate acceptancebetween the users. An important aspect is the coherencethat maintains the avatars and virtual world appearances.Additionally, note that it is required to consider severalalternatives, which should be approved by the stakeholders.For example in the figure the visual effect of changing asimple feature, as is the color environment is also illustrated.For our example the visual concept (b) was selected, thisseems to be most friendly, less traditional and break outtraditional working schemes. Furthermore, these designsmust be integrated with the software development process byusing documentation and formal graphical languages such asUML.

After that visual appearance is defined, the collaborativedesign process should be performed. As was mentionedabove, it is the base for developing interaction and com-munication processes. Therefore, at the stage  (collabora-

Character sheet

(a)Character sheet

(b)

Figure 3. 3D conceptual design

tion model) we propose to appropriate previous conceptsand methodologies that have shown be useful in CSCWsystem designs. In particular, in this example we use theCOoperative MOdel notation based on UML, COMO-UM,which are a modification of the well-known state diagramsof UML 1.5. Details of that notation can be found in [23].shows the collaboration models designed for our system, in(a) the organization diagram and in (b) the collaborative task“Key Finding”. In these diagrams the different user roles andthe interactions between them can be observed, in (a) it ispossible see roles transition by time out event, and (b) showsthe role actions in the collaborative task.

The stage à refers the development of the visualizationand the interaction models. These results from the applica-tion of the collaboration model to the 3D environment basedon the visual parameters arose in the conceptual model. Thisstage also involves the functional and visual GUI design.Figure 5 shows the design of a virtual room in the 3d world,an avatar model. Note that these models are coherent tothe conceptual design selected in Á step. The GUI designmust be include two different schemes, the first one to theactive gamer and the second one for de passive gamer (theobserver). The awareness elements involve the animation ofthe iconic avatar to the passive gamer, which will be usedfor indicating that this is writing a message, audio alertsfor indicating that a message has been sent, visual alerts forindicating to active gamer that the passive gamer requiresyour attention, among others. In this example some conceptssuch as aura, focus and nimbus were not used because theusers not share a common space.

Figure 6 illustrate the interaction model with interactionsdiagrams for each role, in this example we use the Con-

Role Observer1

Role Explorer3

[ T

ime

out]

[ T

ime

out]

Organization

(a)

ExplorerOpen the box

ExplorerBox finding

ObserverObserve to explorer

ObseverGuide to explorer

Protocol conversational -Chat requirements

Knowplace?

Timeout

Iskey?

Task key Fin

yes

no

yes

no

yes

no

Identifya box

yesno

(b)

Figure 4. Collaboration model. (a) Organization diagram. (b) Task diagram

Figure 5. Final 3D designs, example room and avatar

curTaskTrees (CTT) notation [28], that is commonly usedin user interface specification. This model involves a chatsystem but also non-verbal schemes such as attention callingand other awareness elements. As it is expected that thisCVE be executed in corporative environments, we use anavigation pattern that is well known and used in severalvideo games, described in the IV. This is characterizedbecause only uses mouse and keyboard as input devices.

Once design stages have been performed follows thesoftware development stage (Ä). This stage starts with thedesign of a low level architecture that it is then taken toa programming language for generating a software appli-

*

Key finding

Walk Update view

*

Box finding

Chat

AlarmWrite

Translate to other room

orChange role

Clue solving

Read

Observe

[>

||| ||| ||| |||

|||

|||

Sond Update iconicavatar

|||

(a)*

Guide

Sendsparkle

Box finding

Chat

AlarmWrite

Change observed room

Clue solving

Read

||| ||| |||

||| |||

Observe

Change role

Sond Update iconicavatar

|||

[> [>*

(b)

Figure 6. Interaction diagrams. (a) Explorer role. (b) Obsever role

cation. In this stage is also important to define a logicrepresentation model of the 3D elements. In this case, weuse a graph representation of the scene, which make easierthe navigation process, the animation of the avatars and thedetection of possible collisions.

The proposed framework defines two last stages dedicatedto evaluating the system. The first one named collaborationmonitoring (step Å) is focused in to evaluate how the teamperforms the communication, coordination and sharing infor-mation processes. In this case, some metrics such as numberof sent messages, mean answer time, number of attentioncalls, mean time for founding each key, effectiveness of thecollaboration process developed by the passive gamer, whichcould be measured doing an analysis of the system log. Thesecond one Æ ) should evaluate the user immersion basedon the monitoring of the awareness elements.

Interaction Action

↑ Avatar moving at the front

↓ Avatar moving at the back

← Avatar turning at the left

→ Avatar turning at the right

Í Avatar viewing in mouse direction

Table INAVIGATION PATTERN

Finally, the cyclic model of the proposed frameworksuggest that several adjusts could be applied on a specificsystem element such as the visual characteristics, awarenesselements, collaboration model, etc., but also could be appliedto the global system according to an analysis process.

V. CONCLUSIONS AND FUTURE WORK

In this paper a general framework for developing CSCWsystem in VE have been proposed. The main aspect ofthis framework is the fact that requirements specificationand evaluation processes are centered in observation of useractivities when developing collaborative tasks.

The proposed framework is defined as an iterative pro-cess composed by 7 stepsCSCW and 3D software designmodels. The requirements specification has been divided intwo independent processes: software requirements and 3Ddesign. On the other hand, the system modeling is composedof three steps called collaboration model, visualization andinteraction model and software development. These threemodels are integrated for generating a final computationalapplication that ideally achieves all requirements of thesystem. Finally, evaluation process has been divided in twomonitoring processes that attempt to assess the main aspectsof this kind of systems i.e. collaboration and awarenesscapabilities. Due that this framework has been proposed asan iterative process, problems detected in the first iterationwill be fixed in the next iteration process.

Future work will be focused in the definition of detailedartifacts for developing each step of the proposed framework.In that way, we will be interested in the definition of animplementable instance of the proposed framework and theevaluation in some specific collaborative applications.

ACKNOWLEDGMENT

Wilson Sarmiento is fully supported by the UniversidadMilitar Nueva Granada at Colombia.

REFERENCES

[1] P. Ray, “Computer Supported Cooperative Work (CSCW),” inCooperative Management of Enterprise Networks, M. Malek,Ed. Springer US, 2002, pp. 27–46.

[2] J. Menon, “Collaborative visualization and modeling,” inShape Modeling and Applications, Proceedings. InternationalConference on, 1997, pp. 178–187.

[3] E. F. Churchill and D. Snowdon, “Collaborative virtual en-vironments: An introductory review of issues and systems,”Virtual Reality, vol. 3, no. 1, pp. 3–15, 1998.

[4] S. Benford, J. Bowers, L. E. Fahlen, J. Mariani, and T. Rod-den, “Supporting Cooperative Work in Virtual Environments,”The Computer Journal, vol. 37, no. 8, pp. 653–668, 1994.

[5] R. L. Gomes, G. J. Hoyos-rivera, J.-p. P. Courtiat, and R. L.Comes, “Collaborative virtual environments: going beyondvirtual reality,” in ICME ’03. Multimedia and Expo, 2003International Conference on . Proceedings., vol. 2, 2003, pp.II–105–8.

[6] K. Y. Dai, S. S. Zhang, and Q. Fu, “Real-time 3D modelsharing conference system based on agent and VRML,” inSystems, Man, and Cybernetics, IEEE International Confer-ence on, vol. 5, 2001, pp. 3355–3360.

[7] P. Fuhrer, G. K. Mostefaoui, and J. Pasquier-Rocha, “MaDVi-World: a software framework for massively distributed virtualworlds,” Software - Practice and Experience, vol. 32, no. 7,pp. 645–668, Jun. 2002.

[8] S. Otmane, N. Ouramdane-Djerah, and M. Mallem, “Towardsa Collaborative 3D Interaction Model for Cooperative Designin Virtual Environments,” in Poceedings od the 11th Interna-tional Conference on Computer Supported Cooperative Workin Design, 2007, pp. 198–203.

[9] H. Hrimech and F. Merienne, “Interaction and evaluation toolsfor collaborative virtual environment,” International Journalon Interactive Design and Manufacturing, vol. 4, no. 3, pp.149–156, 2010.

[10] C. Carlsson and L. E. Fahlen, “Integrated CSCW tools withina shared 3D virtual environment (abstract),” in Proceedings ofthe INTERACT’93 and CHI’93 conference on Human factorsin computing systems, ser. CHI’93. ACM, 1993, p. 513.

[11] C. Greenhalgh and S. Benford, “MASSIVE: a collabora-tive virtual environment for teleconferencing,” ACM TOCHI,vol. 2, no. 3, pp. 239–261, 1995.

[12] R. C. Waters, D. B. Anderson, J. Barrus, D. Brogan, M. Casey,S. Mckeown, T. Nitta, I. Sterns, and W. Yerazunis, “DiamondPark and Spline: Social virtual reality with 3D animation,spoken interaction, and runtime modifiability,” Presence:Teleoperators & Virtual Environments, vol. 6, no. 4, p. 461,1997.

[13] S. Benford, D. Snowdon, A. Colebourne, J. O’Brien, andT. Rodden, “Informing the design of collaborative virtualenvironments,” in Proceedings of the international ACMSIGGROUP conference on Supporting group work: the inte-gration challenge, ser. GROUP’97. ACM, 1997, pp. 71–80.

[14] O. Hagsand, “Interactive multiuser VEs in the DIVE system,”Multimedia, IEEE, vol. 3, no. 1, pp. 30–39, 1996.

[15] C. Greenhalgh and S. Benford, “MASSIVE: a distributedvirtual reality system incorporating spatial trading,” in Dis-tributed Computing Systems, 1995., Proceedings of the 15thInternational Conference on, 1995, pp. 27–34.

[16] S. Mischke, F. Goetz, R. Hinn, and T. Hampel, “sTeam3D:bringing together virtual communities and CSCW,” in Pro-ceedings of the twelfth international conference on 3D webtechnology. ACM, 2007, pp. 169–172.

[17] A. De Lucia, R. Francese, I. Passero, and G. Tortora,“SLMeeting: supporting collaborative work in Second Life,”in Proceedings of the working conference on Advanced visualinterfaces. ACM, 2008, pp. 301–304.

[18] P. Fuhrer and J. Pasquier-Rocha, “The MaDViWorld Project:An Attempt to Apply a Collaborative Virtual World Paradigmto the Internet,” in Proceedings of 9th World Multi-Conferenceon Systemics, Cybernetics and Informatics, N. Callaos,W. Lesso, and M. Palesi, Eds., vol. IV, IIIS. IIIS, 2005,pp. 148–153.

[19] V. Anupam and C. Bajaj, “SHASTRA-an architecture fordevelopment of collaborative applications,” in Enabling Tech-nologies: Infrastructure for Collaborative Enterprises. Pro-ceedings., Second Workshop on, 1993, pp. 155–166.

[20] O. Stiemerling and A. B. Cremers, “The use of cooperationscenarios in the design and evaluation of a CSCW system,”Software Engineering, IEEE Transactions on, vol. 24, no. 12,pp. 1171–1181, 1998.

[21] R. Iqbal, A. James, and R. Gatward, “Designing with ethnog-raphy: An integrative approach to design,” Advanced Engi-neering Informatics, vol. 19, no. 2, pp. 81–92, 2005.

[22] J. L. Garrido, M. Gea, M. Noguera, M. Gonzalez, and J. A.Ibanez, “Una Propuesta Arquitectonica para el Desarrollo deAplicaciones Colaborativas,” in Interaccion 2004. AIPO,2004, pp. 164–171.

[23] M. Noguera, M. V. Hurtado, M. L. Rodrıguez, L. Chung,and J. L. Garrido, “Ontology-driven analysis of UML-basedcollaborative processes using OWL-DL and CPN,” Science ofComputer Programming, vol. 75, no. 8, pp. 726–760, 2010.

[24] S. Rabin, Introduction To Game Development (Game Devel-opment). Charles River Media, Inc., 2005.

[25] G. W. Lecky-Thompson, Video Game Design Revealed.Course Technolgy, 2007.

[26] S. S. S. Visamsetty, P. Bhattacharjee, and N. Polys, “Designpatterns in X3D toolkits,” in Proceedings of the 13th inter-national symposium on 3D web technology, ser. Web3D’08,vol. 1, no. 212. ACM, 2008, pp. 101–104.

[27] C. A. Collazos, L. A. Guerrero, J. A. Pino, and S. F. Ochoa,“A method for evaluating computer-supported collaborativelearning processes,” International Journal of Computer Ap-plications in Technology, vol. 19, no. 3/4, p. 151, May 2004.

[28] F. Paterno, C. Mancini, and S. Meniconi, “ConcurTaskTrees:A Diagrammatic Notation for Specifying Task Models,” inINTERACT ’97 Proceedings of the IFIP TC13 InterantionalConference on Human-Computer Interaction, Jul. 1997, pp.362–369.