The 2001 annual ASIST conference program was billed as being "extensive", and indeed it was! The 7 days of events comprised 65 sessions, 9 invited panels, 49 contributed papers, 34 posters, 7 receptions, 5 continuing education courses, 2 symposia, and 2 keynote addresses. The sessions and panels included a strong industry presence -- a major strength of the ASIST annual meeting, as the application of theory to practice (and the potential for its application) are emphasized throughout the conference.
The ASIST SIGs were exceptionally active -- of the 65 total sessions, 38 were SIG programs. The SIGs cover a broad range of interests in information science, ranging from international information issues, to knowledge management, to classification research. The fledgling DL SIG has only been active for a little over a year, but it is building a reputation for its energy and support for the DL community. The extraordinary efforts of the DL SIG was recognized in honors earned by two of its officers: Allison Kopczynski, SIG-DL 2001 Chair, received the Year 2001 James Cretsos Leadership Award in recognition of her outstanding leadership in professional ASIST activities; and Suzie Allard, the 2002 SIG-DL Chair, received the Year 2001 SIG Member-of-the-Year Award in recognition of her significant contributions to the membership of the SIG through participation in and support of its meetings, publications and other activities.
This year the conference boasted two innovations: the introduction of poster sessions, and the provision of digital recordings of all sessions. The poster sessions were highly popular, both with presenters and with attendees. The session recordings will be made available to attendees either on CDROM or by Internet access. It was noted in one of the sessions that the restriction of these documents to paying attendees illustrates the difficulties of balancing the demands of commercial publishing and the desire of many Internet collection builders to provide free, open access to information.
The two keynote addresses were placed to "bookend" the conference -- appearing as the first and the final conference sessions. The initial keynote was given by Brewster Kahle, inventor of the Internet Archive and currently president of Alexa Internet. His talk centered around the idea that libraries -- conventional as well as digital -- should support how people remember, learn, and create. He noted that support for "remembering" through archiving is fairly strong, although long-term preservation is still problematic; while the Internet Archive believes in "preservation by replication", Kahle notes that at present the IA has achieved, at most, two copies of documents in the IA collection. The IA has itself experienced data loss, much of it caused by programmer errors (according to Kahle, "one person's optimization is another's data loss").
The problem of supporting how people learn with a digital library was revisited in several later sessions, particularly those dealing with knowledge management. The industry panel sessions emphasized the importance of "growing" intellectual capital, as well as recording and preserving it. The focus is on making contributions to the corporate knowledge management system -- in digital library terms, the library's document collection -- a natural and integral part of every organizational function. The key appears to be creating significant incentives for an organization's members to add to the knowledge base, and to ensure that user needs drive the effort of knowledge/document collection and organization.
Another recurring theme of the conference was the need for user studies, to discover what users want or need and how they prefer to access information. The Information Needs, Seeking and Use SIG (SIG USE) was particularly active in organizing sessions. In understanding users and their needs, a focus should be placed on identifying groups of users with common goals and needs. These user communities can be tapped to drive a digital library -- to guide in the development of the initial document base, to use (and learn from) the collection, to create new additions to the digital library through their own contributions. The Internet Archive, for example, is currently developing more focused collections -- an election collection, a September 11, collection, etc.
Supporting a particular community of users through a digital library or knowledge base also implies supporting personalization of interface and search/browsing results for that user group. A generic "one size fits all" library is less likely to achieve significant user support. Case studies from industry illustrated the critical importance of both communicating with potential users before creating a digital library/knowledge management system, and of planning a user-focused deployment strategy for the system. System developers were also urged to bear in mind the heterogenicity of the identified communities -- to avoid the problem of creating a "mass customization" that does not actually satisfy the wants and needs of any particular individual in the target group.
While personalization was emphasized, many presentations also discussed the pressing need for greater standardization in digital library/knowledge management system development tools. As one speaker lamented, "there are a lot of people out there who sell snake oil in the morning and automatic taxonomy generating software in the afternoon."
Collection developers naturally wish to measure or assess the effectiveness of their system with the target user community. Several commonly used techniques were discussed: usability studies, "good citizen metrics" for contributors to the collection (in a commercial organization, sometimes accompanied by a carrot-and-stick approach to encouraging quality contributions), and transaction log analysis. A multi-faceted approach is likely to give the best picture of the community's information needs and the best ways to support them -- and, of course, a community is a dynamic entity, and so these studies should be on-going.
The final keynote presentation, a debate between Dr. James Hendler and Dr. Ben Shneiderman, was indeed as lively as it was billed to be. While James Hendler argued for the development and deployment of adaptive autonomous agents in the near future, Ben Shneiderman adopted a more cautious approach. Shneiderman believes "trust is what we have in people, not machines"; agent technology is not yet ready for prime time; and a principled set of empirical testing and user studies should precede the release of any software. Ironically, although Hendler argues for autonomous agents, he did admit to driving a car with a manual transmission because he believes he can manage shifting better than a machine -- which makes one wonder how well he believes a software agent can handle comparatively more complex tasks. Nevertheless, Hendler scored a point on the topic of careful preliminary user studies when he pointed out that "people are still doing careful empirical studies of hypertext, and they're showing that the Web won't work!".
(On December 18, 2001, the subtitle for the ASIST Meeting was corrected from "Harvesting the Flow" to "Harnessing the Flow".)
Copyright 2001 Sally Jo Cunningham