An Awareness Provisioning Methodology for Asynchronous Virtual Global Forums
Farhad Daneshgar, Elijah van der Kwast, University of New South Wales
Virtual Global Forums (VGF) are web-based asynchronous discussion forums where discussions are organized according to topic and sub-topic. VGF is based on the conversation social mechanism with its own specific limitations when compared to other collaboration support systems that are based on awareness and coordination social mechanisms. VGF currently provide members with added capabilities of getting involved in various simulated virtual collaborative tasks using mediated technology, and ability of users to selectively reveal themselves. While the users favor all these capabilities, these capabilities can also pose challenging barriers to the users’ ability in sharing knowledge among them when compared with similar collaborative activities in both physical as well as synchronous virtual worlds; the reason being a lack of mechanism for maintaining awareness of the users on the VGF. This paper introduces a conceptual framework for (i) representation of the awareness requirements of the actors in VGF processes, and (ii) introducing a measure for such requirements. A case study is used to demonstrate application of the proposed framework to a football fan VGF.��
Virtual Global Forums are advanced form of the traditional structured messaging systems that started to attract much attention of the early researchers in the field of Computer-Supported Cooperative Work (CSCW) in late 1980s (Malone T, et al, 1993) & (Borenstein N, Thyberg C A, 1993). Also called electronic bulletin boards,� discussion forums, and web-based discussion boards, VGF are situated within the different-time/different-place cell of the time/place groupware taxonomy. They are Internet web sites that use the Internet technology embedded within HTML. Like newsgroups, VGF have discussions that are organised according to topic and subtopic where users can participate. Unlike newsgroups however, VGF are not catalogued as part of the public Usenet service on the Internet. The Usenet search engine, therefore, does not search what is written in a forum.�
Both in the past and today, VGF have been used for facilitating informal, unplanned social interactions among geographically dispersed people in a structured manner in asynchronous mode, and are already very popular for both recreational use (Lenhart et al, 2001), as well as in the workplaces (Nardi et al, 2000). Recent versions of VGF allow users (that is, forum members, administrators, guests/visitors and moderators) to interact with multimedia objects in creative ways. These versions also allow a diverse set of virtual collaborations among their users, which is the subject matter of this paper.
Virtual collaboration in this paper is referred to a set of activities that connects people, share information among people, facilitates information gathering and use, and facilitates cooperative work and education among people. Some examples of virtual collaboration in VGF include various discussions that lead to some kind of virtual collaboration in the form of betting, voting, shopping, and banking where the unit of economic trade is some hypothetical gold bullions accumulated in the members’ virtual bank accounts as a result of reaching various ranks among the community by say, length of membership period, good manners, winning in hypothetical lotteries and donations by other members. All such collaborations are conducted through exchange of ideas and knowledge under various discussion threads for an example of VGF visit the http://www.persianfootball.com .
With few exceptions, awareness has been regarded by CSCW (Computer-Supported Cooperative Work) researchers as a “kind of information that is highly relevant to a specific actor/role and situation of a process participant” (Baker et al 2002). On the other hand exists the concept ‘knowledge-sharing’ that refers to the seamless transfer of knowledge to individuals (Turban & Aronson, 2001, p.361). Rooted in the above arguments is an awareness-based methodology for enhancing knowledge sharing capabilities in virtual communities called the ‘process awareness framework’ (Daneshgar, 2004a), (Daneshgar et al, 2004b). The aim of this methodology and its underlying framework is: (i) to introduce a representation tool for virtual collaborative processes, (ii) to provide a measure for knowledge-sharing capabilities for members of virtual communities in general, and (iii) to enhance collaboration in these communities.
This paper introduces a specialised version of the above framework for virtual collaboration in VGF by taking into consideration specific attributes of VGF processes as described below. A case study is also used to evaluate suitability of the framework for identification of the awareness and knowledge sharing requirements of the users of VGF. It is expected that the results of this paper guide design of VGF groupware systems that, among other things, facilitate virtual collaboration among their users.��
2.�������� Knowledge Sharing In VGF
Like many social clubs in the physical world it is also common for VGF to target certain slice of population such as parents, teenagers, and football fans, and perhaps expecting members to register before entering certain transactions within the forum. The difference however is that the true identity of members in VGF may not be validated as accurately as in their physical counterparts. Related to this is another major difference between the two types of communities: an additional ability of the members of VGF to selectively hide themselves from others. In fact a human member, referred to as an actor in this paper, can enact multiple personalities within the VGF without being detected as easily as in the physical world.�
While a great deal of research has already been conducted in providing presence awareness in instant messaging systems that are synchronous, not much work has been done on presence awareness for asynchronous systems such as VGF. Some of the earlier attempts in providing presence awareness in synchronous systems include high bandwidth and information rich presence awareness through live video feeds. Refinements of this kind of presence system were made to provide lower bandwidth iconic presence indicators instead (Greenberg, 1996). One recent system is BuddySpace. It is an advanced synchronous instant messaging client that aims to reduce or eliminate many of the drawbacks of instant messaging through enhanced presence awareness and management (Eisenstadt et al, 2003). The key idea behind the BuddySpace is to incorporate geographic maps that can display the location and availability of various contacts. Study is also in progress in the Centre for Advanced Technologies at AMS on presence awareness and management features and how they could be implemented in an instant messaging client. Other commercial instant messaging systems are AIM and ICQ. They integrate limited text-based presence indicators such as “Online” and “Idle”. Also, the Babble system provides a graphical display of active different individuals who are in text conversations (Erickson et al., 1999). As mentioned before, and as far as the writers are aware not much research has been conducted for maintaining awareness of the actors in (asynchronous) VGF other than iconic symbols for presence. Such limitations may impose undesirable restrictions on the kind of collaboration that VGF users can get involved in; a collaboration that is currently simulated through exchange of messages.�
Generally speaking, in any collaborative process the ability to share knowledge is an essential requirement of effective collaboration. It is argued in this paper that modelling techniques used for collaborative processes need to possess the ability to accurately address and represent knowledge-sharing requirements of its actors. Hawryszkiewycz (1997) asserts that quality processes require the ability to retain memory about decisions that were made during the course of a process (i.e., group memory). However in order to maintain group memory systematically, the process must be able to identify/represent the way knowledge is shared among the actors in the first place.
When performing virtual collaborative activities on VGF there is a strong relationship between the actors’ awareness of other actors and their activities in one hand, and the concepts of focus and nimbus as used by social psychologists, on the other hand. These concepts are subsets within which an object chooses to direct either its presence or its attention. “The more an object is within your focus, the more aware you are of the object and the more an object is within your nimbus, the more aware the object is of you” (Benford et al, 1994). As a result, the awareness levels of actors in VGF can be conceptually calculated by combining several nimbi and foci related to other actors, their tasks, and the artefacts that they use in order to perform their tasks. Actors in VGF require certain level of awareness about various aspects of virtual collaboration before they can perform their tasks within the forum. For example, an actor will not be able to get involved in a ‘lottery’ activity unless s/he knows about a sale of lotteries within the forum, of who is initiating the sale, and how lottery tickets can be purchased, and how the results are announced. During an ethnographical study authors noticed that despite having sufficient number of gold bullions on hand and willingness to buy lottery tickets within the VGF, no obvious presence awareness for this object (that is, when and where the lotteries were needed) existed on the forum, and individual messages must have been traced back in order to find relevant information for such purchase. In other words, ‘lotteries’ did not produce sufficient nimbi (plural form for nimbus) in order to overlap with the researchers’ focus when and where they were needed most. Also, it was noted that in their current form, individual messages do not trigger artefact creation/modifications; a capability that can greatly enhance collaboration in VGF.�
Initially originated from the field of social science, the term contextual awareness referred to determining what information actors should be made aware of, and how actors should be made aware, thus limiting distractions from the intended context (Liechti 2000). Such contextual knowledge seems to be pre-requisite for the actors’ effective virtual collaboration within the VGF. This approach to knowledge sharing and collaboration is also consistent with the growing view shared by many in the community of knowledge management that demands clear separation of context, narrative and content management with the aim of enhancing collaboration and self-organising capabilities of the knowledge-based communities (Snowden 2002). In the next section a framework is proposed for representation of contextual knowledge in VGF processes with the aim of providing a measure for, controlling of, and enhancing, e-collaboration in VGF.��
3.�������� Representation Of Collaborative Knowledge In VGF
According to the interactionist school in social psychology, awareness is maintained if each person actively provides a kind of nimbus by which s/he selectively exposes some of his/her properties (that is, their activities, etc.) to others. According to this school, pairwiseinteractions between people occur either by nimbus (an object’s presence), or by focus (its attention); the more an object is within one’s focus, the more aware the person is of it. Also, the more an object provides a nimbus the more aware others will be of it (Benford et al, 1994). In VGF, nimbus-related issues such as shyness (desire to hide, and yet to participate), odour (‘iconic faces symbolising the actor’s true identity’, ‘request to collect signature for letter sending to the President’, or ‘willingness to donate gold bullions’) particularly become quite relevant and take different courses when compared to physical communities. In VGF the absolute/fixed knowledge exists within the actors’ minds whereas how much they are willing to share is highly situated on the actor’s. As a result, the same context can be seen differently by different people, depending on both the willingness of the object to expose itself (nimbus), as well as the eyesight of the viewers (focus). In other words, the awareness level of the actors in VGF depends on both their foci (plural for focus) as well as the others’ nimbi (plural form for nimbus). For example, moderators of a VGF may wish to construct their own less public virtual spaces within the VGF in order to discuss confidential issues related to certain aspect of the VGF process such as banning unethical actors/members, censorship issues, making decisions about various standards and policies such as editing certain words/contents and so on. Such knowledge needs to be shared by the VGF moderators only. As far as the focus and nimbus concepts are concerned, arranging such secret meeting in physical world has quite different implications from the same meeting in a VGF: what if one of the moderators him/her self is also the same member with a bad mouth swearing at others?
The above discussion implies that knowledge representation and visualisation in VGF have their specific characteristics, and that specialised frameworks are required in order to represent, measure and enhance collaboration in VGF. The following section introduces one such framework.
3.1.����� An Awareness Framework for Sharing Collaborative Knowledge in VGF
The proposed framework consists of two components. The first component is an ontological knowledge representation graphical model for the VGF process called the awareness net and is explained in 3.1.1. The second component is an awareness model that defines various levels for awareness of the actors in VGF process, and is discussed in 3.1.2.
3.1.1.�� An Awareness Net For VGF
Awareness net is a representation of the contextual knowledge about collaboration aspects of the VGF process, and consists of a collection of semantic concepts linked together according to a set of predefined relationships. Its purpose is to demonstrate a measure for awareness as well as identifying awareness requirements of the actors within the VGF. Its main focus is not on the actors’ behaviour at the interface or the flow and transformation of information within the VGF but mainly on the cooperation among the actors. This is particularly true for VGF where people’s interactions can take many forms depending on the purpose of the interaction or the tasks the actors perform.
Many collaborative process models have already been developed that provide a common representation format that facilitate human understanding and communication, and at the same time support and automate collaborative process management (Nabuco et al, 2001), (Lee et al, 2000), (Barker et al, 2002), (Farquhar & Rice, 2003) with each focusing on a different issue. However none of these models explicitly address the awareness requirements of actors. Awareness net provides a unified analysis of both the knowledge-sharing requirements of the collaborating actors within the VGF process, as well as the VGF process itself. In other words, the same analytical tool is used for representation of both the awareness as well as the collaborative process. Using a connected graph, knowledge-sharing requirements of the actors are identified in terms of the collaborative semantic concepts used for the construction of the VGF process. In order to maintain such unified analytical approach, a rule-based model was developed that constitutes the second component of the proposed framework called the awareness model, and is discussed in 3.1.2.
An awareness net of a VGF for football fans is shown in Figure 1, and consists of the following collaborative semantic concepts:
����������� (i) A non-empty set of role vertices and simple task vertices (as opposed to collaborative task). A collaborative task is a pair of simple tasks each performed by a separate role. Role is a set of norms that are enacted by the actors and are expressed in terms of obligations, privileges, and rights. Roles and actors are used interchangeably in this study.�
����������� (ii) A set of edges or arcs that connect above classes together and represents various artefacts used by roles in order to execute their tasks.��
The edges in the graph represent artefacts that are used by actors when performing various tasks. A separation is made between the two kinds of artefacts. One is called the roleartefact and has a role and a simple task as its endpoints. This artefact, among other things, encapsulates personal knowledge/resources the actor would use to execute the task successfully. In Figure 1 bold ovals represent tasks, filled ovals represent roles, straight lines represent role artefacts and dotted lines represent task artefacts. The other kind of artefact is called task artefact, which has two simple tasks as its endpoints. These two simple tasks together constitute a collaborative task. Two collaborating actors, that is actors with a common task artefact, use/exchange/act-upon their common task artefact in order to collaborate. In order to represent interactions between a pair of instances of the same role class, such as interactions among the VGF common members, a double arrowed line is used that represents both the role artefact and the task artefact that the two instances use for their collaboration. Following Case Study demonstrate awareness net for a VGF for football fans.
CASE STUDY: A Football Discussion Forum
A case study was conducted to demonstrate and analyse the awareness requirements of the actors in VGF. Based on the recommendation by Sarantakos (1998) for the selection of a suitable case, a pilot study was initially conducted to select a VGF with the highest level of collaboration intensity, as measured by foci and nimbi. This pilot study provided a quality source of observer training and contributed to limiting potential bias as a result of insufficient variations among relevant variables. A sample of 20 Sport VGF was investigated. These VGF were classified based on the degree of support that they provide for collaboration among their users. The intention was to select the one with highest score of support for collaboration. The pilot study helped to identify the following set of criteria for collaboration support in relation to focus and nimbus:
�Allowing users to use graphic emoticons in their messages
�Allowing the notification of personal messages via emails of pop-up application windows
�Notification of a posted reply
�Presence of a user profile feature
�Users’ access to other users’ profiles
�Users’ ability to protect their profiles
�Users’ ability of searching for other users
�Can users see which other users are on-line?
�Presence of a topic status feature
�Does the site allow avators?
�Can members log in under an invisible mode?
This pilot study led us to a Sports VGF that possessed most of the above features. This VGF was then investigated with the aim of identifying its various semantic concepts including roles, tasks and artefacts. The fact that this site possesses maximum features of collaboration support in the selected sample will ensure accuracy of results for validating the applicability of the awareness net analysis to the VGF in future studies.� Awareness net analysis was applied to this site and a summary of results is shown in Figure 1.
Figure 1 – An Awareness Net for the Sport VGF
Above awareness net shows three main classes of roles (filled ovals) namely Administrators, Registered Users, and Unregistered Users, each role performing a set of tasks within the community using relevant artefacts. For example A Registered User interacts through topic, with other Registered Users (double arrow line), the subject of their interaction being their ‘topic’. This is an example of a role collaborating with another instance of the role. An example of collaboration between two different roles is when a Registered User creates a message using relevant knowledge for doing so, in collaboration with the Administrator, who has to process such message creation request, using his/her VGF administration knowledge. The message itself is the artefact that is created/used/authorized/published through a sequence of steps, collaboratively by these two roles. And finally, an example of a non-collaborative task is when the Administrator uses his topic knowledge to create a hidden topic. Interestingly, this hidden topic will eventually will be used as a task artefact, by an Administrator for the collaborative task of ‘interaction through hidden topic’ with other Administrator(s). For detailed explanation of the Awareness Net refer to Daneshgar (2004a).�����
3.1.2��� The Awareness Model
Sharing the contextual knowledge in VGF occurs at various levels depending on both the foci as well as the nimbi related to various semantic concepts that make up the VGF process. Five levels are identified for generic VGF and are discussed below:
Level-0 Awareness: An actor is at level-0 if s/he possesses contextual knowledge about the concepts (that is, roles, tasks, role artefacts and task artefacts) that lead the actor to an understanding of the tasks that the actor performs within the VGF. This knowledge is the sum of the actor’s foci (his visibility and eyesight) as well as the tasks’ nimbi (how clearly the tasks are presented to, and conceived by, the actor). Columns 2 and 3 of Table 1 show objects that constitute level-0 awareness for the corresponding roles in column 1.
Table 1 – Level-0 Awareness Objects for the Sport VGF.
Level-1 Awareness: Level-0 awareness is a pre-requisite for the next level up. An actor that reaches level-1 awareness will possess knowledge about all the concepts that leads the actor to awareness about all of the actors within the VGF with whom the actor has a direct task dependency. Columns 2, 3 and 4 in Table 2 show objects/concepts in addition to the level-0 awareness, that make up level-1 awareness for corresponding actors listed in column 1.�
Table 2 – Level-1 Awareness Objects for the Sport VGF.
�Level-2 Awareness: An actor at level-2 awareness will have knowledge about the objects that lead the actor towards an understanding of the remaining roles within the VGF. It is the actor’s level-1 awareness, plus all the objects that are required in order for the actor to know all other actors within the VGF.
Table 3 – Level-2 Awareness Objects for the Sport VGF.
Level-3 Awareness: An actor at level-3 awareness has the knowledge about the objects that lead that actor towards an understanding of all interactions that occur between any pair of roles within the VGF. Attaining level-3 awareness enables an actor to initiate level-3 context-sharing transactions within the VGF. Columns 2, 3 and 4 in Table 4 shows objects in addition to those for level-2, indicating level-3 awareness for the roles in column 1.
Table 4 – Level-3 Awareness Objects for the Sport VGF.
Level-4 Awareness: An actor at level-4 awareness has a specialised knowledge about the objects that lead that actor to an understanding of how all the objects within the VGF. Graphically, this level can be represented by the whole of the awareness net. At this level the actor has adequate contextual knowledge about what everyone does within the VGF, how they perform their tasks (that is, the sum of everybody’s level-0 awarenesses), who directly collaborates with whom and how (sum of everybody’s level-1 awarenesses), and who directly or indirectly collaborates with whom and how (sum of all the level-2 awarenesses), and finally, how other actors collaborate with one another (sum of all the level-3 awarenesses). From the Figure 1 the additional objects that brings up awareness levels of the actors to level-4 are shown in columns 2, 3 and 4 of the Table 5 below.
Table 5 – Level-4 Awareness Objects for the Sport VGF.
4.�������� Conclusion and Future Work
This paper introduced a conceptual framework for representation of contextual knowledge in VGF collaborative processes. The ultimate goal of this representation is to identify awareness requirements of the actors in VGF that in turn may lead to enhancements in collaboration services in VGF (say, by automating some ofthese services). It is expected that as a result of this identification, the administrative staff of VGF be able to provide various levels of collaboration support for their members both more efficiently, by automating such support, and more effectively, by adopting a systematic method for awareness provisioning mechanism for keeping members aware of various aspects of collaboration in VGF. Work is in progress to investigate specific and detail directives for designing groupware systems that maintain awareness requirements of these actors at required levels of awareness at all times. The theoretical aspect of the proposed framework is also intended to improve by developing a conceptual framework for identification of objects that constitute knowledge gaps for various actors within existing VGF processes, and to facilitate awareness provisioning mechanisms for removing such undesirable gaps by bringing into the focus of the actors such required knowledge. Some earlier conceptual studies in the field of CSCW (Computer-Supported Cooperative Work) for achieving similar objectives in physical communities include ‘encouraging articulation work’ (Gerson 1986), ‘explicification of situated actions’ (Allen 1984), ‘identification of mutual influence’ (Robinson 1991a), ‘facilitation of shared views/shared materials’ (Robinson 1991b), and ‘provision of a double-level languagethat allows both ambiguity and clarity’ (Robinson 1991a). Also there exist relevant research tracks in the fields of Knowledge Management, e-collaboration and Virtual Organisations that will constitute foundation for our future research in enhancing collaboration in VGF. Historical developments of these concepts will be investigated in order to address the issue of collaboration enhancements in VGF.
Allen, T. J. (1984). Managing the Flow of Technology. MIT Press,
Barker, D., Georgakopoulos, D., Schuster, H., Cichoki, A. (2002). Awareness Provisioning in Collaboration Management. International Journal of Cooperative Information Systems, Vol. 11, Nos. 1 & 2 (2002), pp. 145-173.
Benford, S., Bowers, J., Fahlen, L., Mariani, J., Rodden, T. (1994) Supporting cooperative work in virtual environments, The Computer Journal, vol. 37, no. 8, pp. 653-668.
Borenstein, N.S., Thyberg, C.A. (1993), “Power, ease of use and cooperative work in a practical multimedia message system”, Readings in Groupware and Computer-Supported Cooperative Work, edited by Ronald M Baecker, Morgan Kaufmann Publishers, Inc. USA, pp.485-500.
Daneshgar, F. (2004a) A Methodology for Sharing Contextual Knowledge in Virtual Communities. Chapter in Innovations in Knowledge Management� by B. Montano, Idea Group Publishing, .
Daneshgar, F., Ray, P., Rahbi, F., Molli, H.S., Molli, P., Godart, C. (2004b); “Knowledge Sharing Infrastructures for Teams within Virtual Communities”, in e-Collaborations and Virtual Organizations, edited by Dr. Michelle Fong, IGP/Infosci/IRM Press, Hershey, PA, USA, 2004.
Eisenstadt, M., Komzak, J., Dzbor, M. (2003) Instant messaging + maps=powerful collaboration tool for distance learning. In Proceedings of TelEduc03, May 19-21.
Erickson, T., Smith, D.N., Kellogg, W.A., Laff, M., Richards,� J.T., Bradner, E. (1999) Socially Translucent Systems: Social Proxies, Persistent Conversation and the design of “Babble”. In Proceedings of ACM CHI 99 Conference on Human Factors in Computing Systems, May 15-20.
Farquhar, A., Rice, J. (2003), The Ontolingua Server: a tool for collaborative ontology construction. Stanford’s Knowledge Systems Laboratory, ; http://ontolingua.stanford.edu
Gerson, E., Star, S. (1986). Analysing due process in the workplace, , ACM Transactions on Office Information Systems 4(3).
Greenberg, S. (1996) Peepholes: Low Cost Awareness of One’s Community. In Proceedings of ACM CHI 96 Conference on Human Factors in Computing Systems, April 13-18.
Hawryszkiewycz, (1997). Designing the networked enterprises. Archer House inc. , .
Lee, J.D., Hickey, A.A., Zhang, D., Santanen, E., Zhou, L. (2000), “CoID SPA: A Tool For Collaborative Process Model Development”, 33rd Haeaii International Conference on System Sciences, Hawaii, USA.
Lenhart, A., Raine, L., Lewis, O. (2001) Teenage Life Online: The rise of the instant message generation and the Internet’s impact on friendships and family relationships. Pew Internet & American Life Project, , Retrieved from http://www.pewinternet.org/reports/toc.asp?Report=36
Liechti, O. (2000) Awareness and WWW: An overview, Sigroup Bulletin, vol. 21, no. 3, pp. 3-12.
Malone, T.W., Grant, K.R., Lai, K.Y., Rao, R., Rosenblit, D.A. (1993), “The Information Lens: An Intelligent System for Information Sharing and Coordination”, Readings in Groupware and Computer-Supported Cooperative Work, edited by Ronald M Baecker, Morgan Kaufmann Publishers, Inc. USA, pp.461-473.
Nabuco, O., Drira, K., Dantas, E.(2001), “A Layered Design Model for Knowledge and Information Sharing Cooperative Systems”, IEEE 10th International Workshop on Enabling Technologies, Cambridge, MA, USA.
Nardi, B., Whittaker, S., Bradner, E. (2000) International Outeraction: Instant Messaging in Action. In Proceedings of the ACM Conference on Computer Supported Cooperative Work, 79-88.
Robinson, M. (1991a). Double level languages and cooperative working. AI & Society, 5:34-60.
Robinson, M. (1991b), “Computer Supported Cooperative Work: Cases and Concepts”, Proceedings of Groupware 91, SERC, P O Box 424, 3500 AK Utrecht, the Netherlands.
Sarantakos, S. (1998) Social Research, 2nd Ed., Macmillan Publishers, Melbourne, Australia.
Snowden, D. (2002) Complex acts of knowing: Paradox and descriptive self-awareness, Journal of Knowledge Management, vol.6, no.2.
Turban, E., Aronson, J.E. (2001), Decision Support Systems and Intelligent Systems, Prentice Hall, NJ, USA.
Contact The Authors: Farhad Daneshgar and Elijah van der Kwast
Farhad Daneshgar, PhD, Senior Lecturer, School of Information Systems, Technology and Management, University of New South Wales, Sydney, Australia, Address: F.Daneshgar, SISTM, UNSW SYDNEY 2052 AUSTRALIA Email: firstname.lastname@example.org; Phone: +61 (2) 9385 4241; Fax: +61 (2) 9662 4061 Web Page (click on Academic Staff, Farhad Daneshgar): http://www.sistm.unsw.edu.au/