Lee Li-Jen Chen and Brian R. Gaines
Knowledge Science Institute
University of Calgary
Calgary, Alberta, Canada T2N 1N4
The work reported in this article is part of a program of research at the Knowledge Science Institute concerned with the systematic acceleration of scientific research through the use of information technology to support collaboration in distributed research communities. One of the problems noted in the development of systems such as Mediator (Gaines and Norrie, 1994) to support international collaboration in manufacturing and manufacturing research was that of maintaining awareness between remote partners that activities had occurred in one location that affected those in another. The implementation of Mediator on the World-Wide Web (Gaines, Norrie and Lapsley, 1995) makes provision for collaborative access to distributed files, including workflow models, but does not provide mechanisms to draw attention to changes in those files or updates to those models. CHRONO, a chronological indexing systems described in this article, was designed to support improved awareness of change by those using Mediator. The development of CHRONO has also raised more general issues about the support of awareness in communities operating through the web, and this article develops a general framework for these issues and the existing tools designed to support awareness.
The World Wide Web was originally conceived and developed at CERN for the purpose of assisting and facilitating collaborative interactions among high energy physicists, working at various institutions in different countries, to conduct joint research projects. The idea of the web was prompted by experience of a hypermedia system used for keeping track of personal information on a distributed project, and developed into something with much greater connotations:-
The World Wide Web (W3) was developed to be a pool of human knowledge, which would allow collaborators in remote sites to share their ideas and all aspects of a common project. (Berners-Lee, Cailliau, Luotonen, Nielsen and Secret, 1994)The phenomenal growth of the web can be attributed to its emergent growth property: the ability for a new state of being to emerge naturally from a synergy among existing systems. The web is structured such that if it was used independently for two projects, and later relationships are found between the projects, then no major or centralized changes have to be made, but the information can be linked to represent the new state of knowledge. This property of emergent growth has allowed the web to expand rapidly from its origins at CERN across the Internet irrespective of national or disciplinary boundaries.
The dynamics of the web are based on three fundamental notions: (i) computer-supported cooperative work; (ii) hypermedia; and (iii) cross-boundaries, emergent growth. The implications of combining these three basic notions to support the web's raison d'Ítre (that is, to be a pool of human knowledge) will be explored in the following sections. A common thread through the subsequent discussion is the concept of awareness. Group awareness is essential to provide smooth coordination among members in a collaborative project team. Through community awareness, the web as a global hypermedia system supports an emergent CyberSociety (Jones, 1995) that transcends the traditional boundaries of both physical and social communities.
This article is concerned with modeling the psycho-social dynamics of the Word-Wide Web in relation to computer-supported cooperative work (CSCW). In the subsequent sections, awareness issues in collaborative group are discussed, an experimental system for facilitating chronological awareness on the web is described and a preliminary conceptual framework is developed. This framework is used to analyze existing chronological awareness tools and general indexing systems on the web in terms of Smith's (1994) collective intelligence model and Miller's (1978) living systems theory. A preliminary sketch of a CyberOrganic conceptual framework for the web is presented and explored.
This section presents a brief survey of CSCW research in respect to awareness issues, and describes a taxonomy of different types of awareness involved in group collaboration.
Most CSCW research shares the goal of trying to assist people to work collaboratively as a cohesive team and to provide them with a sense of common purposes (e.g., completion of the group task). For example, Landow's (1990) In Memoriam project utilizes hypertext's freedom of navigation and linking ability to break down physical separation and the univocal voice of textual conversation. In so doing it creates a new awareness of the processes of collaborative learning and collaborative work for group members in literary studies. Ishii and colleagues (Ishii and Miyake, 1991; Ishii and Kobayashi, 1992) have created a fusion of video and computer workspace to provide a seamless working environment with remote workstations, in order to provide mutual awareness between distance collaborators in real-time. Olson and Atkins' (1990) NSF EXPRESS Project uses intelligent, multimedia email to facilitate cooperation within scientific and engineering community by increasing researchers' awareness about each other's work.
A critical requirement in shared tasks is maintaining situational awareness (Norman, 1993) by keeping everyone adequately informed. In an environment where each member has a well defined role, the need to have face-to-face communication in order to perform a cooperative task becomes less necessary if mechanisms for situational awareness have been well established between members. For example, the navigation of a large ship requires effective coordination of various people with differing roles (Hutchins, 1990). Many key members of the navigation team (conning officer, plotter, bearing takers, deck log keeper, bearing-timer recorder and fathometer operator) are geographically separated (pilot house, chart house, and port and starboard wings) and communicate with each other by a common telephone circuit. The common audio channel and the physical layout of the pilot house provide opportunities for the navigation crews to observe each other's work, contributing to partial redundancy in their joint knowledge. They also support maintenance of the group over time to provide fault-tolerance if some group members fail to perform their roles.
Thus, one of the important criteria for achieving group cohesiveness is not the opportunity of being in constant face-to-face communication, but rather the situational awareness of what other group members are doing. It can be achieved through either face-to-face (as on the pilot house) or telephone communication (as between the port and starboard wings). Together the functional specificity of the crews and the cognitive artifacts that facilitate situational awareness (e.g. the single telephone circuit, the pilot house with high visibility among navigational team members) can create an effective collaborative system.
The emphasis on the importance of social interaction and cognitive artifacts (such as the telephone circuit described previously or the web) as the means to enhance human abilities has also been echoed by Norman (1991). A cognitive artifact is defined as "an artificial device designed to maintain, display, or operate upon information in order to serve a representational function." The power of a cognitive artifact comes from its function as a representational device. However, artifacts do not actually change an individual's capabilities. Rather, they change the nature of the task performed by the person. When the informational and processing structure of the artifact is combined with the task and the informational and processing structure of the human, the result is to expand and enhance cognitive capabilities of the total system of human, artifact, and task.
Hence, when designing a computer-supported collaborative system CSCW researchers need to keep the representational aspect of the whole system in mind. A system designed with this holistic perspective in mind can provide tools that enhance communication, coordination and social interaction capabilities. Therefore, the collaborator's intentions can be easily transferred to each other and common goals can be achieved with mutual satisfaction. In the various studies described above, situational awareness has acted as an important coordination mechanism among collaborative groups.
Recently, Smith (1994) has proposed a collective intelligence theoretical model that provides a holistic perspective on the nature of computer-supported intellectual collaboration. It focuses primarily on project groups and/or relatively small, close-knit teams within larger groups in designing/creating artifacts (e.g., committee reports, computer software). He calls such a group-artifact-task arrangement, artifact-based collaboration (or ABC). His comprehensive model takes into account cognitive processes within individuals, artifacts to-be-built, and situated actions.
The following sub-sections describe a taxonomy of group awareness (what Smith calls collective awareness) based on the collective intelligence model. Using human self-awareness model and cognitive Information Processing System (IPS) architecture (Newell and Simon, 1972; Newell, 1982) as references, two analogous forms of awareness can be identified for collective groups: awareness of the group's collective long-term memory (LTM) and awareness among and of each other.
Global Awareness -- The most general level is the body of intangible knowledge that is shared by all member of the groups. It includes the overall goals of the project, its ways of operating, the strategies it uses to develop the artifact, its current status and problems, the relation of the project to the external environment, etc. This awareness is not deep, if the project is large, but it provides each member with a sense of the whole.
Deep Awareness -- At the other extreme is the deep, detailed, often technical, knowledge held by individual members. Depending on the project, a single individual is often responsible for a particular part. Thus, the level of awareness and expertise required to generate a segment of the overall artifact (project) is significantly greater than that required for another person to understand it.
Peripheral Awareness -- Between the extreme of general, shared knowledge and deep, individually specialized and generative knowledge, is an intermediate level: the thick knowledge of adjacent or near specialty areas. It takes the form of understanding, rather than generation. Thus, it is shared with individuals or the team responsible for developing other parts of the artifact, but it is not as deep as their knowledge nor is it shared with the entire project. This critical peripheral awareness is ultimately responsible for the integrity of the group's work. It provides a context for the interfaces between areas. Thick shared knowledge can be developed through informal interactions, such as conversations, but it can also be developed through more formal mechanisms, such as institutionalized reviews (e.g., structural walk through process in a software project).
Thus, the model identifies three forms of awareness with respect to a group's LTM:-
A different kind of awareness is the awareness members have of one another. This category of awareness is closer to the notion of situational awareness described earlier.
Resource Awareness -- One of the primary reasons for assembling a group is to assemble the expertise required to carry out a project. Therefore, the issue becomes one of providing the group (as a whole) with a collective awareness of its members' respective specialized knowledge and expertise. Thus an extremely valuable resource for a group is shared knowledge of who is an expert on what.
Task-Socio Awareness -- Another form of awareness involves the interaction between social and intellectual processes operating within in the group. It would be simple if groups were purely intellectual organisms, but they are not: tensions exist and factions develop. These developments are inevitable. For example, one member may oppose an idea voiced by another not because the idea is objectionable but because of who said it. The opposite condition -- support an idea because of friendship or attraction -- is equally bad. These so very human situations are unlikely to go away, but a group should be aware of itself as a dynamic, functioning organism as well as be aware of the artifact it is developing to insure that the integrity of its work is not compromised by them.
Chronological Awareness -- A third form of awareness is the instantaneous awareness that an individual has regarding the activities of other individuals. This is what constitutes chronological awareness. For example, one member of the group may be aware (or wish to know) what another member of the group is working on in a nearby part of the artifact. This behavior is monitored at a very low level by the collaboration support system in its concurrency control mechanisms to insure that two members do not attempt to change the same part of the artifact at the same time. These mechanisms, however, do not prevent one member's access from blocking that of another, or prevent one member's subsequent work from affecting earlier work done by another. Groups may need help in monitoring domains of activity. For example, members may want to see where colleagues are working; they may even wish to see a display over time of the "tracks" left by colleagues.
Thus, the model identifies three forms of awareness among members of one another:-
This section examines various issues relating to awareness of changes in a working environment. The focus here is on chronological awareness, that is, the awareness of when something (an event or an artifact) has changed. Members of a group may be interested to know someone has modified an artifact, or when a new (and potentially relevant) artifact has been created.
In a dynamic environment where large amounts of information are created and updated frequently, the need to keep up with the most up-to-date and relevant information has become more important as the web's community expands. In terms of group collaboration, to be aware of changes is one of the fundamental requirements for coordination and for providing a sense of connectedness.
Some of the characteristics of human intellectual work that are valued most highly are: coherence, consistency, correctness, and elegance. It is difficult to imagine how work with these attributes could be produced without that structure of ideas having been held in its entirety by a single mind -- if not actually produced by that mind (Smith, 1994). As stated previously, however, by considering awareness from a functional point of view, one may be able to construct mechanisms that enable groups to achieve comparable result. Web pages (in addition to FTP archives, Listservers, etc.) are becoming the primary means for information dissemination, and these web pages are being constantly updated to reflect members' current states of knowledge on their portions of collective memory.
As a part of the collective awareness taxonomy, a chronological awareness support system for the web provides each individual an instantaneous awareness of other individuals' activities. Hence it allows team members to synchronize their activities in a more coherent way by keeping them informed and aware of any changes made to each other's web pages that might be relevant to their current tasks.
In addition, a chronological awareness support system saves time and effort in several ways: first, there is no need for group members to sit in front of their web browsers (e.g., Mosaic, Netscape) clicking and checking each other's document separately; and, second, web users do not need to check every URL they monitor just to learn that nothing has changed. The system indicates for the users the web documents that have changed and need their personal attention.
CHRONOis an HTTPD server-side system which generates chronological listings of web pages that have been changed recently at specific sites. It provides a basic awareness-support that lets visitors to a web site (e.g., members of a group, an organization or other Netsurfers) see which web pages have been modified since their last visit. Currently, the CHRONO system is implemented on a UNIX platform. CHRONO presents to the visitors an HTML document that lists the titles of web pages at the site in reverse chronological order (Figure 1). This chronological listing of web pages also functions as a collection of hyperlinks to the listed pages.
Figure 1 Screen Snapshot of CHRONO
The user interface of the CHRONO system is fairly straightforward and intuitive for web users. It looks like an automated what's new page to the users. From the list, the visitors will be able to tell at a glance what documents have been modified or created recently. They can also scroll down the list to check those older documents at the site. Because the titles of the listed pages also act as hyperlinks to the actual web pages, the visitors can simply click on them and jump to the relevant pages of interests.
Therefore in addition to typical hyperlinks in HTML documents which provide linkage to related information, chronological hyperlinks presented in the CHRONO listings provide the visitors the means to access the newly modified or created pages. This time-line (or history) dimensionality of a chronological listing complements the functionality of the associative memory characteristic found in typical hyperlinks that join related information.
This time-line dimension allows frequent visitors of a web site an immediate awareness on what have been changed since their latest visit. The changes may reflect some web pages in which they have been previously interested or they may show some pages that they have never seen before but now appeal to them. Hence this chronological browsing characteristic is analog to spatial (subject-category) browsing characteristic that library patrons have often experienced when looking for books on open book-shelves (i.e., accidentally finding (more) relevant books near the books that they were looking for originally).
The difference is that instead of finding relevant information via browsing the near by subject-categories, the users may now find relevant information via browsing the concurrently modified/created web pages. Sometimes, conceptually related documents are created (or modified) around the same time, however their author(s) may not remember to update the HTML links to them. Unlike a manually updated what's new page in which the users have to rely on the timely updates made by a webmaster (or by the document authors), CHRONO provides the time-line dimension to the users automatically, in a reliable and periodic fashion.
The current experimental version of the CHRONO system is available only on the UNIX platform, but should be easily ported to other operating systems. It is written in the C language and uses the standard UNIX file I/O library to traverse through the directories of a targeted web site.
The system consists of a background process that is invoked by a UNIX system daemon cron on an hourly base. When invoked, this process recursively enters any pre-specified directory that comprises a web site. It looks for HTML files, then extracts their title field and records their time/date of last modification. If a document does not have a title field, its file name is used instead. The process sorts the collected titles into reverse chronological order. Finally it produces a HTML document containing the listing. The result is a master web document containing the hyperlinks to these periodically generated chronological listings. A new listing page is generated via the background process for its targeted web sites every hour.
Currently the system has no mechanisms for recording when a particular user last visited the site, and relies on the users' own recollection. Future versions of the system will incorporate the ability to provide customized information to the users.
The CHRONO system has been set up for the Department of Computer Science at the University of Calgary since March, 1995. Because this is still an experimental service, its services have been limited five web locations: two research units and three individuals. People associated with the two research units: Knowledge Science Institutes and GroupLab periodically have utilized the system to check on new developments of each other (both within group and between groups).
From a preliminary examination of the HTTPD access_log of the CHRONO system site and from talking with individual group members, we have found that the system indeed has provided chronological awareness for group members at both the group and the organization levels. Most people have found such chronological awareness-support service helpful to them in keeping themselves informed about works of others. Occasionally, there have been visitors from other institutions browsing the chronological listings, but such accesses have been infrequent, the service has only recently been made generally known.
The chronological listings of the three personal sites have offered other group members more focused chronological awareness about these three individuals' working patterns. Occasionally, some people had discovered new projects the targeted individuals were working on that they were not previously aware. This positive usage experience of personal chronological listings suggests to us that there is a need to further examine the effectiveness of providing different sub-groupings of chronological listings to group members. For example, chronological listings for particular sub-directories of a web site, for particular projects of a group, or for chronological access pattern of particular documents may prove to be useful for different occasions and user needs. The last example (i.e., the chronological access pattern) can be categorized as an instance for supporting mutual awareness.
In summary, the current experience of the CHRONO system has indeed demonstrated its usefulness for frequent visitors of a particular web site in formulating chronological awareness of changes at the site. It has been used by group members of two research units keeping tracks of new developments between the groups and within the group. Finally, there are different ways of sub-grouping chronological listings at any given site. They may prove to be useful for different purposes and under different circumstances. Therefore they require further investigation to determine how they function as cognitive artifacts in supporting chronological awareness.
There are three other systems that provide support for chronological awareness based on different designs and implementations. These are briefly examined in this subsection, followed by a comparative evaluation of current chronological awareness systems.
WebWatch is a client-side chronological awareness system for keeping track of changes in selected web documents. Given an HTML document referencing URLs on the web, it produces a filtered list, containing only those URLs that have been modified since a given time (Figure 2).
Figure 2 Screen Snapshot of WebWatch
The chronological criteria used for filtering can be given as a global setting that applies to all URLs, or can be derived automatically, using the time of user's last visit to the document, as recorded by the web browser in the user's local HTML (bookmark) file.
The system generates a local HTML document that contains links to only those documents which were updated after the given date. Later on, the user can load this document into any web browser and use it to navigate to the updated documents.
In contrast with the simple line listing strategy used in CHRONO, it stores its arguments in a parameter file. Once the users have customized the program to their needs, using its graphical front-end, they can have it run periodically in unattended mode.
WebWatch retrieves only the "header" of a document, to check its "Last Modified" date. The size of this header is usually quite small compared to that of the entire document. Hence, the user connect time, and ultimately the load on the network, is significantly reduced.
URL-minder is a centralized system that keeps track of resources on the web and sends registered users e-mail whenever their personally registered resources change. Web users can have the URL-minder keep track of any web resource accessible via HTTP. It can be anything -- not just web pages users personally maintain (Figure 3).
Figure 3 URL-Minder Fill-in Form
The chronological awareness system keeps track of one web page, image file, or other resource at a time. It will not keep track of all the web pages linked to the page that a user submits. A separate URL must be submitted for every distinct page users want the URL-minder to track for them. The URL-minder tracks the actual HTML markup, binary contents, or ASCII contents of the URL they have submitted. If an HTML page includes a GIF or JPEG graphic, the URL-minder will inform them when the reference to the graphic changes. If the users want to know when the actual content of a binary graphic file changes, they must submit the URL to the binary graphic itself. The URL-minder currently checks on users' registered URL's at least once per week, and will inform them if it fails to retrieve their registered URL after trying twice.
This subsection presents a comparative evaluation of the four chronological awareness support systems discussed previously: CHRONO, WebWatch, Katipo, and URL-minder. Each system has its unique approaches for achieving chronological awareness support for web users and complement each other along three main dimensions:
The first dimension, the locus of responsibility, differentiates who is responsible for maintaining the record-keeping mechanisms for chronological awareness. For example, CHRONO is a server-side system in which chronological listings are being updated and kept at the web server-side. This strategy ensures that only the users who are current visiting the web site would need to know what information has been changed. Hence, it reduces network traffics by avoiding needless broadcasting of chronological information to some users who might not be concerned with it. Thus, CHRONO can be thought of as offering "chronological awareness on demand". The main disadvantage of this strategy is that in order to know whether or not any particular page has been changed, a user would need to check out the specific web site periodically. WebWatch and Katipo, however, put the responsibility of maintaining chronological awareness on the client-side. Both client side systems periodically monitor specific pages at various web sites and report whether or not they have been changed recently. Hence, both ensure the user would be aware of any changes. The main disadvantage is that the users need to remember to run such a system, or it must be set up to run periodically, in turn causing it to consume higher network bandwidth. Finally, URL-minder requires its users to register at a centralized site, so that it can automatically monitor the registered pages for the users. Its main disadvantage is the high network traffic involved in such a centralized broadcasting scheme.
The second dimension, the method of locating changes involves two different ways of locating documents that have been changed: browsing and targeting. CHRONO uses the browsing approach in order to facilitate the chronological browsing characteristic: visitors of a site may find relevant information via browsing the concurrently created/modified web pages, because closely related documents are sometimes created (or modified) around the same time. This browsing approach allows the at glance attribute for accidental discovery of relevant information without prior awareness of their existence. Conversely, WebWatch, Katipo, and URL-minder employ a targeting approach in which they are targeted on specific pages or information that users have previously specified. Therefore this method of locating changes is more direct and efficient. However the main disadvantage of such approach is that the users cannot be made aware of any new information which have been created recently. They are limited only to changes made to prior knowledge.
Finally, the third dimension, the complexity of user interaction, denotes system usability in terms of simplicity vs. customization. CHRONO and URL-minder are in the simplicity category in the sense that their user interfaces are simple and familiar to web users (i.e., scrolling list of hyperlinks and fill-in form of URLs and e-mail address). They are geared toward ease of use and a shallow learning curve. Both systems, however, have no capability for individual customization. In contrast WebWatch and Katipo allow elaborate customization of features, but they also demand more efforts by the users to learn and utilize their functionalities.
Therefore, each chronological awareness support system examined so far have various degrees of advantages and disadvantages along the three dimensions. CHRONO has the advantages of: (i) simplicity of user interface; (ii) supporting accidental discovery via its browsing characteristic; and (iii) server-side chronological awareness information on demand. It is nicely complimented by WebWatch and Katipo for their strength in the efficiency of targeting approach and customization capabilities. And finally URL-minder offers another unique service: it uses e-mail as its notification channel. This approach is useful for users who use their e-mail systems more frequently than web browsers. Therefore together as a whole, these chronological awareness support systems have covered a wide range of approaches in respect to three major dimensions of chronological awareness support.
Communities (or groups, organizations) are open systems. That is, a community maintains a relatively stable structure and boundary while receiving inputs from the environment, processing them, and extruding outputs. The human components of organizations -- individuals and groups -- are also open systems. Furthermore, these open systems are composed primarily of living entities -- cells, organs, and organisms (Tracy, 1989).
Living systems theory is the life's work of James Grier Miller (1978). The basic systems theory was built upon a search for the common properties of all living systems. Miller's theory of living systems goes beyond general and open systems theories in describing the vital systems and processes inherent in cells, organs, organisms, groups, organizations, communities, societies and supranationals (Figure 4).
Figure 4 Levels of Living Systems and Nonliving Environment
Viewed from this open systems perspective, the web and its user communities can be considered as the integral parts of a cybernetic living system at the meta-community level. It is natural, therefore that living systems theory should be applied to the understanding of this emergent CyberOrganism and its psycho-social behaviours. Such an ensemble from the human-machine-environment triad can be conceptualized as a cross-over between CyberSpace (Benedikt, 1991) and Virtual Communities (Jones, 1995). The CyberOrganic context is mainly focused on the four levels: organism (i.e. individual), group, organization, community of living systems theory (the shaded areas in Figure 4).
This subsection briefly describes the basic concepts in living systems theory (LST) as a foundation of a basic overall conceptual framework for the web. The basic living systems theory was built upon a search for the common properties of all living systems. It should be noted at this point that the basic theory is in flux, and may continue to be (Bailey, 1994). LST demonstrates that living systems exist at eight levels of increasing complexity from cells, organisms, to groups, communities, and supranationals.
Originally, the book Living Systems (Miller, 1978) presented 19 basic subsystems at seven levels, and since then, James Grier Miller and his long time collaborator and co-author Jessie L. Miller, have added a 20th subsystem, the timer, and eighth level, the community (Miller and Miller, 1990; Miller and Miller, 1992). The community level was added between the organization and the society (Figure 4). The timer subsystem was added to the original list of nine information-processing subsystems, making a total of 10 (not counting the boundary and reproducer, which process both information and matter-energy). The twenty subsystems (shown in Figure 5) are responsible for the ongoing day-to-day operation of the living system, it is these subsystems which keep the system alive (Bailey, 1994).
Figure 5 The 20 Critical Subsystems of a Living System
The survival and health of individuals, groups, organizations, communities, and societies depend on performance and coordination of a set of essential processes. In all, LST identifies twenty critical subsystems carrying out these processes at every level. For example, a channel and net subsystem to convey information from one part of the system to another was found to be necessary in cells as well as societies. The nature of that subsystem might vary but the function remained the same (Tracy, 1989). Some critical subsystems process matter-energy, some process information, and some process both. From this holistic perspective, the chronological awareness for community can be viewed to be one of the fundamental components (in the intersection of the timer critical sub-system and the community level) in the living systems theory. A comprehensive presentation of the critical sub-systems in the LST, is given by Miller (1978) and Tracy (1989). For the purposes of this article the timer sub-system is most relevant to chronological awareness.
A Timer subsystem transmits information to a decider subsystem about time-related environmental states or about of components of the system. This information signals the deciders of subsystems to start, stop, alter the rate, or advance or delay the phase of one or more of the system's processes, to coordinate them in time. The timer consists of one or more oscillators known as clocks or pacemakers, the phase of which can be reset. They measure duration or order in time or underlying rhythms of various sorts. The timer synchronizes internal processes of the system and coordinates the system with its environment (Miller and Miller, 1990). For example, at the group level, a mother wakens other family members on time; at the organization level, workers take regular monthly inventory; and at the community level, artifacts such as clocks mark the opening and closing of schools and buildings, as well as regulating traffic lights and parking meters. The same can also be applied to a larger time scale, in terms of annual community celebrations of local and national holidays.
Smith's (1994) Collective Intelligence Model can be characterized as a cognitive model within the living systems theory. It focuses on two system levels: the individual (organism) and the group levels (the lightly shaded inner areas in Figure 4). The model regards collaborative groups as a form of information processing system, analogous to Newell's and Simon's IPS model of individual cognition (Newell and Simon, 1972; Newell, 1982).
Within the model, a collective memory system includes subsystems that provide a collective long-term memory (LTM) for tangible knowledge, built and maintained in a computer system, and for intangible knowledge, carried in the heads of the human beings that comprise the group. The memory system also includes working memory (the currently attenuated awareness) for both types of information.
A collective processor includes the fine-grain operations used by groups to develop, access, and maintain the information stored in the memory system. The collective processor as a whole can be viewed as a loosely coupled distributed system that includes multiple independent processors, joined by communications and social networks (Smith, 1994).
Collective strategy enables coherence in collaborative work. Individual processes occur not in isolation but in purposeful sequences. These strings of operations are analogous to statements in a language intended to accomplish a goal or to communicate a message. The system responsible for generating sequences of operations is analogous to the grammar individuals used to generate a string of words. Therefore a collection of collective strategies become a work-flow model for the collaborative group.
Finally, the model considers two meta-cognitive issues: collective awareness and collective control. Many collaborative projects are too large and too complex to be understood by any one person. However, people often expect groups to produce work with the same integrity and consistency sometimes found in work produced by a single good mind working alone. By developing thick, overlapping areas of shared knowledge, groups may be able to piece together a form of collective, but distributed peripheral awareness that is sufficiently coherent to achieved this goal. Control must also be distributed over a group. Otherwise, information will not flow across boundaries, and the group and its work will be brittle. However, although many decisions can, and probably made by consensus, authority must ultimately be centralized in order to resolve disagreements and to preserve the integrity of the group's work (Smith, 1994). Here chronological awareness provides essential information for groups to exercise distributed, collective controls that maintain smooth coordination.
When the collective intelligence model is viewed with in the context of the living systems theory (LST), its basic concepts of human-computer symbiosis, collective awareness, and collective control can be extended to formulate a broader conceptual framework for the World Wide Web. This LST-oriented conceptual framework characterizes the web as an emergent living system that is called CyberOrganism.
Within in this CyberOrganism perspective, individuals, groups, organizations, and communities together create a collective intelligence that becomes a pool of human knowledge for individual collaborators in remote sites to share their ideas and all aspects of collaboration. In order for the CyberOrganism to function in a cohesive and integrative manner, coordination at various system levels must be maintained. However the means for maintaining web cohesion and integrity are still in a gradual process of emerging and evolving.
One of the main issues involved in achieving system coordination is the notion of awareness maintenance. Awareness maintenance is concerned about how subsystems in various levels of the CyberOrganism (namely, group, organization, and community) be aware of each other, both within specific levels and between them. Such maintenance of mutual awareness is essential whenever individuals, groups or organizations are involved collaborative tasks.
The previous section has presented chronological awareness support systems as a type of systems designed mainly to support awareness maintenance for individuals at the group level (and to some degree at the organization level). Currently there are other types of awareness maintenance systems on the web that address different system levels. The following section examines and categorizes them along two dimensions: level of awareness and locus of responsibility.
The previous section has discussed the concept of system levels in the living systems theory, and described a CyberOrganism framework for the web that focuses on the dynamic interactions between individuals, groups, organizations and communities. Within this framework, individuals constitute the fundamental level of analysis in the overall system. Self awareness provides a sense of identity, purpose, and consciousness to each individual. When perceiving all levels collectively, the notion of self awareness can be extended to different hierarchical partitions in the system for the purpose of analyzing the inter-relationship between the overall system and its parts in the collective stance (Gaines, 1994).
Hence, collective awareness occurs at every level above the individual level in the system hierarchy. In order to function as a coherent whole, mechanisms for maintaining awareness in the group, organization and community levels will need to exist first. Currently there are various tools and systems on the web that have emerged to fill the roles of providing awareness maintenance for various levels.
The web uses the client-server model to partition the computational division of labor. Similarly the locus of responsibility of awareness maintenance at every level can be divided into originators of information and retrievers of information. An originator is a source/server of information dissemination and an retriever is an user/client of information resource. Within the living systems theory, originators and retrievers of information are situated at the opposite ends of the channel and net information subsystems (Miller, 1978). The various awareness maintenance mechanisms (such as the chronological awareness support systems for groups) serve as (or provide the functionalities of) the timer, associator, and memory subsystems in the CyberOrganism.
Therefore using the two dimensions: level of awareness and locus of responsibility, a taxonomy of web tools and systems for awareness maintenance can be outlined.
Using the locus of responsibility dimension at the group, organization and community level, one can categorize various tools and systems for maintaining situational awareness of the current information resource on the web (Figure 6).
Figure 6 Taxonomy of Mechanisms Supporting Awareness Maintenance
At the group level, originators of the information resource can organize and implement work flow models of the group activities, use server-side chronological awareness support systems such as CHRONO, or/and send e-mail notification to users of information. Retrievers of the group level information resource can use client-side chronological awareness tools such as WebWatch and Katipo or register in centralized dispatcher service like URL-minder. Alternatively, they can send e-mail to inquire to information originators to see if any new things have come up.
At the organization level, originators of the organizational resource can broadcast to concerned individuals, groups, or organizations via specific Listservers, register in NCSA's What's New service, or announce in organization-maintained MOO or MUD (such as JHM). They can also establish what's new HTML links in organization news while retrievers can participate in HyperNews or MOO, and follow the new HTML links in organizational web pages.
At the community level, originators can register the information resource in hierarchical subject services like Yahoo , or initialize their pages in searching and navigational services like Web Crawler while retrievers can browse Yahoo, search LYCOS or Web Crawler, or read and post the USENET groups.
The previous sections have developed an overall conceptual framework based on living systems theory and the collective intelligence model. Such a collective stance toward modeling the web as a living system presents a basic framework for structural-functional descriptions of the emerging CyberOrganism; that is, the perspective of the web as a constantly evolving phenomenon/artifact that emerged from the dynamic interactions between the community organization, group, individual partitions in an open-system.
The article first introduced the issues involved with various aspects of awareness in CSCW. This collective awareness taxonomy provides a useful starting point toward an overall conceptual framework for the web. In this article, the focus has been on chronological awareness, and an experimental awareness-support system has been described. Finally, the last section explored how the chronological awareness system together with various existing systems on the web can fit into the CyberOrganism framework via two dimensions of awareness maintenance: level of awareness and locus of responsibility. This provides a taxonomy of web tools and systems for awareness maintenance.
The conceptual framework provides research directions to further investigate the psycho-socio nature of the web. For an example, in term of chronological awareness, there is a need to further examine the effectiveness of providing different sub-groupings of chronological listings to group members, such as chronological listings for particular sub-directories of a web site, for particular projects of a group, or for chronological access patterns of particular documents which may prove to be useful on different occasions and for varying user needs.
In addition, the issues of establishing mutual awareness between individuals, groups, organizations need to be explored further. Also, the various critical subsystems of the CyberOrganism have to be identified and analyzed more critically. Currently one can tentatively identify some parts of this emerging living system, however the dynamics of interactions between various critical subsystems are vaguely understood and require further exploration.
One future research direction is to conduct focused studies on specific communities (e.g., various specialized fields in scientific/scholarly disciplines) that utilize the web as a major part of their integrative working environment. Using the conceptual framework stated so far as a guide and observational methodologies for group dynamics (e.g., SYMLOG, Bales and Cohen, 1979; Losada, Sanchez and Noble, 1990), it will be possible to investigate further the psycho-socio dynamics of those communities on the web.
This work was funded in part by the Natural Sciences and Engineering Research Council of Canada. We are grateful to Wayne Robbins for his careful reading and correction of an earlier draft.
Bailey, K.D. (1994) Sociology and the New Systems Theory: Toward a Theoretical Synthesis. New York, Free Press.
Bales, R.F. and Cohen, S.P. (1979) SYMLOG: A system for the multiple level observation of groups. New York, Free Press.
Benedikt, M., Ed. (1991) CyberSpace: First Steps. Cambridge, MA, MIT Press.
Berners-Lee, T., Cailliau, R., Luotonen, A., Nielsen, H.F. and Secret, A. (1994) The World-Wide Web. Communications ACM 37(8) 76-82.
Gaines, B.R. (1994) The collective stance in modeling expertise in individuals and organizations. International Journal of Expert Systems 7(1) 21-51.
Gaines, B.R. and Norrie, D.H. (1994) Mediator: information and knowledge management for the virtual factory. SIGMAN AAAI-94 Workshop: Reasoning about the Shop Floor. pp.30-39. Menlo Park, California, AAAI.
Gaines, B.R., Norrie, D.H. and Lapsley, A.Z. (1995) Mediator: an Intelligent Information System Supporting the Virtual Manufacturing Enterprise. Proceedings of 1995 IEEE International Conference on Systems, Man and Cybernetics. New York, IEEE.
Hutchins (1990) The technology of team navigation. Galegher, J., Kraut, R.E. and Egido, C., Ed. Intellectual Teamwork: Social and Technological Foundations of Cooperative Work. pp.407-428. Hillsdale, New Jersey, Erlbaum.
Ishii, H. and Kobayashi, M. (1992) ClearBoard: a seamless medium for shared drawing and conversion with eye contact. Proceedings, CHI'92. New York, ACM.
Ishii, H. and Miyake, N. (1991) Toward an open shared workspace. Communications ACM 34(12) 37-50.
Jones, S. G. (1995) Understanding community in the information age. Jones, S. G., Ed. CyberSociety: Computer-Mediated Communication and Community. pp.10-35. Thousand Oaks, CA, Sage.
Landow, G.P. (1990) Hypertext and collaborative work: the example of Intermedia. Galegher, J., Kraut, R.E. and Egido, C., Ed. Intellectual Teamwork: Social and Technological Foundations of Cooperative Work. pp.407-428. Hillsdale, New Jersey, Erlbaum.
Losada, M., Sanchez, P. and Noble, E.E. (1990) Collaborative technology and group process feedback: their impact on interactive sequences in meetings. CSCW'90 pp.53-64. New York, ACM.
Miller, J.G. (1978) Living Systems. New York, McGraw Hill.
Miller, J. G. and Miller, J. L. (1990) Introduction: the nature of living systems. Behavioral Science 35 157-196.
Miller, J. G. and Miller, J. L. (1992) Greater than the sum of its parts I. Subsystems which process both matter-energy and information. Behavioral Science 37 1-38.
Newell, A. (1982) The knowledge Level. Artificial Intelligence 18(1) 87-127.
Newell, A. and Simon, H.A. (1972) Human Problem Solving. Englewood Cliffs, NJ, Prentice-Hall.
Norman, D. A. (1991) Cognitive artifacts. Carroll, J. M., Ed. Designing Interaction: Psychology at the Human-Computer Interface. pp.17-38. Cambridge, UK, Cambridge University Press.
Norman, D. A. (1993) Things That Make Us Smart: Defending Human Attributes in the Age of the Machine. Reading, MA, Addison-Wesley.
Olson, G.M. and Atkins, D.E. (1990) Supporting collaboration with advanced multimedia electronic mail: the NSF EXPRES project. Galegher, J., Kraut, R.E. and Egido, C., Ed. Intellectual Teamwork: Social and Technological Foundations of Cooperative Work. pp.429-451. Hillsdale, New Jersey, Erlbaum.
Smith, J.B. (1994) Collective Intelligence in Computer-Based Collaboration. Hillsdale, New Jersey, Erlbaum.
Tracy, L. (1989) The Living Organization: Systems of Behavior. New York, Praeger.
Last update: 2002-03-27 by Lee Chen