The opening session will be an interactive brainstorming event in which we will attempt to define a set of critical issues facing communities and organizations in rapidly changing times; these issues will form the foci for the following sessions. This first session will be a time to share our experiences to date in addressing these issues.
We will discuss the consequences of today's incessant technological change using examples from the computer field: the changing nature of the Internet and the phenomenon of Java, for instance. How can computer developers and computer users keep pace with ubiquitous transmutation of their domain?
We will review the frameworks we have developed and explored in the recent past in our efforts to get a handle on change. Among the central concepts are the notion of LifeLong Learning as on-going, task-driven learning-on-demand and the idea that computer-supported knowledge bases need to develop through cycles of seeding, evolution and reseeding. Driving our system-building efforts has been the conviction that computational medialike the design environments suggested by our various prototypescan help people function in complex, information-intensive settings more effectively than can traditional paper-based media.
Presentations from L3D members will illustrate our approaches to organizational learning and task-driven learning with the Evolving Artifact (EVA) system and the Chart N'Art learning environment.
As life becomes exponentially more complex and the world shrinks around us, it becomes increasingly clear that learning and knowing are social acts -- that they typically take place within organizational contexts that motivate, reward and influence the learning. Moreover, the boundaries between personal knowledge and organizational memories is fading away; many professionals must make knowledgeable use of more information than any individual can digest and keep up-to-date with. Organizations depend upon new employees having access to knowledge constructed in the past by people who may no longer be available. Organizational memories are needed that transcend the skulls of individuals. However, traditional computer databanks lack the affordances of personal memories. And the Web is no panacea as a repository of world-wide knowledge -- it is far too noisy, disorganized and unreliable.
We need to develop new paradigms of computational group memories that can meet the needs of flexible organizations and changing information sources. These group memories need to have features like those that made personal memories usable and useful. They need structures that are meaningful, helpful, comprehensible and natural so that people can find information they need even when they are not aware of what they need or that they need it.
Presentations from L3D members will show examples of organizational memories that capture ideas and allow group members to share, review and use those ideas. The first prototype system, GIMME, illustrates mechanisms for capturing email discussions and indexing them without relying on a fixed, predetermined set of keywords. Another groupware prototype, Collaborative Info Environment (CIE), features mechanisms for a reseeding process of merging multiple versions of evolving documents.
We conceptualize the social groupings within which domain knowledge evolves as "communities of practice." These are organizations, institutions, professional groups, networks of people or specific cultures that share a body of domain knowledge and that transform themselves as they evolve their knowledge base. Individual newcomers and successive generations of practitioners alter the functioning of the community as they construct their own roles, novel contributions and self-understandings. Their learning is situated in the very context that is redefined and evolved by their learning.
Computer support for such learning must be tailored to the ways communities of practice communicate, learn, evolve. For instance, it must provide for differences in how individuals view the shared knowledge, as well as allowing for constant evolution of the knowledge base itself. This evolution takes place at many levels: facts are accrued quantitatively, but the qualitative categories and significance of the facts changes as well. To be useful, systems must start out with a reasonable base of knowledge (a seed which is designed for growth); they must be structured to accommodate unforeseen additions (evolution); and they must permit periodic restructuring (reseeding) to prune unnecessary duplication, to reorganize confused relationships and to extend functionality to meet newly arisen needs.
Presentations from L3D members will describe systems we are developing to explore evolution of knowledge within two types of communities: computer network designers and elementary school students. WebNet is an attempt to move our previous design environment ideas to the Web, where they can be shared by a potentially large virtual community as well as by specific local communities of practice. WebQuest is a game creation environment in which students design adventure quests that incorporate finding factual subject matter on the Web, so that the students construct learning environments for their peers.
The final session of the series will focus on concrete mechanisms for achieving the goals that arose in the preceding sessions. Here we will try to collect and discuss ideas for our future research: What are some effective forms for group memories? How can they be made computationally powerful? What is the role of digital libraries for various communities of practice? What kinds of information retrieval do we need and what kinds do we already have that might be put to better advantage?
This session will provide an opportunity to discuss mechanisms seen in the demos, approaches described in the lead-in talks to the sessions, and other ideas from people's work or from the research literature. We can explore questions of the relationship between our theoretical frameworks and these implementation mechanisms. We can also raise issues of how to evaluate evolution processes and the effectiveness of computer support for them.
Presentations from L3D members will discuss mechanisms for sharing information over the Web. The Agentsheets Remote Explorium allows people to download games and other simulations: work is underway to promote sharing of individual simulation components like specific agent behaviors. The Teacher's Curriculum Assistant helps teachers to locate , adapt and share curriculum, allowing globally distributed resources to be personalized to the needs of local classrooms.
Go to top of this page
Return to Gerry Stahl's Home Page
Send email to Gerry.Stahl@drexel.edu
This page last modified on August 01, 2003