Preprints of the Metadiversity Conference Proceedings

  Session 4: Building the Infrastructure

Digital Libraries Research and Infrastructure

STEPHEN GRIFFIN, Program Director, Division of Information and Intelligent Systems at the National Science Foundation (NSF)

I would like to begin by telling you that I am from the National Science Foundation (NSF). The NSF is a government agency. Frequently people get us mixed up with the National Academy of Sciences, so I just wanted to say that we are an Executive agency, very much in the same scheme of governmental management as other independent agencies.

The organizational structure of NSF is that of a common bureaucracy. Bureaucracies are easy to laugh about. But they can seriously get in the way of getting things done, particularly in interdisciplinary programs. That is something about which we have thought very hard regarding the Digital Libraries Initiative (DLI), for not only is it interdisciplinary, it is inter-everything. How does one, therefore, map support for an inter-everything set of activities onto a very tightly structured hierarchical set of organizations–a set of super-organizations–that are established purportedly to promote these sorts of things?

The Digital Libraries Initiative is the focus of my talk today. We have two phases–Phase 1 and Phase 2, and I would like to present them both.

The Digital Libraries Initiative – Phase 1

Phase I officially ended at the end of August, 1998, although most of the sites still have their testbeds and their Web pages up online. It was a fairly modest program cosponsored by the National Science Foundation, the Defense Advancement Research Projects Agency (DARPA), and the National Aeronautics and Space Administration (NASA) to try to put together some technology testbeds and some programs of basic research. Make no mistake, the Digital Libraries Initiative was an information technology program. We provided no money for content.

The research that was involved in the Digital Libraries Initiative projects could be characterized as 90% digital, 10% libraries (that did not escape everyone's notice). We think that we had some very successful technology projects. Over the course of the four-year term of DLI-1, as our thinking matured and evolved, we came to the conclusion that developing digital libraries technologies needed to proceed within consideration of multiple contexts, including content, social, domain, applications, and international. A technology-centric program would not be as successful. We feel that the DLI-Phase 1 projects were exceptionally successful in adopting and working within broader perspectives.

We have put in $24 million over four years, and the cost sharing varied between 100% and 200% per project. So government investment was effectively doubled and tripled.

There were six projects, all based on a similar project model and each led by one of the following universities: the University of California at Berkley, Stanford University, the University of California, Santa Barbara (whose testbed was closest and most germane to geographic information systems), the University of Illinois, the University of Michigan, and Carnegie Mellon University. We managed to broaden the scope of the projects, increase the level of government investment, and actually further our thinking about what we were doing by building partnerships with people who had other values and other insights and other perspectives on the idea of a digital library. I think we had, in all, over 100 partnerships of many different types. Some corporations would give funds to the projects. Some would supply services or other resources, including staff.

Some of the most productive forms of partnerships were in the exchange of staff. Some of the companies paid for researchers from their research laboratories to spend a year at a university. Even in a globally networked world, there are some things that can only be achieved when people are face to face, sitting together around the same table. So, I think this exchange of personnel was extremely valuable. We hope to expand this concept and start exchange programs with visiting scientists.

The Digital Libraries Initiative – Phase 2

In February of last year, based on the recognized success and achievements of the Phase-1 projects and other community efforts, we put together a second phase to the initiative, a phase we creatively called Digital Libraries Initiative-Phase 2. In this phase we broadened the program in a number of ways, particularly in terms of content use and usability. Once again the technology only gains meaning when it is placed in an environment where people use it to work on information of value and interest to them. The problems associated with that makes it very difficult to, for example, persuade the computer science community that it should use some of its efforts and resources to assist other communities. Believe it or not, I cannot make, without some scrutiny, an award to a biologist. This is because it would be noted in this community, where people keep very close track as to whether your program is funding the people they believe it is supposed to fund. So one of the real challenges to the Digital Libraries Initiative-Phase 2 is to construct new funding schemes that will desensitize this issue.

Partnerships for Phase 2

The core sponsors of the Digital Libraries Initiative-Phase 1, NSF, DARPA, and NASA, are joined in Phase 2 by the National Library of Medicine, the Library of Congress, the National Endowment for the Humanities, and the FBI. We have partnerships established with a number of other agencies, including the Smithsonian, the National Archives, and the Institute for Museum and Library Services.

I believe strongly that digital libraries research and development activities must be carried out in places other than university laboratories. We must have activities that we support in venues other than university research and computer-science laboratories. We must have activities supported in museums and in libraries, wherever they may be. As a result of this conviction, our new program made very substantial changes in the documentation. In Phase 1, competition was open to universities. In Phase 2, this competition is open to both universities and non-profits, thus multiplying the number of eligible proposers by a significant amount. We also are trying to make the Digital Libraries-Phase 2 attract a much broader sponsorship across the federal government and the private sector by trying to encourage other people to join in the funding of projects that come out of our competition.

Narrowing the Technology Agenda

In addition, the technology agenda is narrowed. We have discovered during Phase 1 that certain technological areas are more important to the digital libraries than others. So, we have a narrower technology focus. We are pushing extremely hard to develop and to enhance research activities related to developing content and collections. We are also pushing hard within our organization to find digital libraries infrastructure.

Phase 2 Programs

Digital Libraries-Phase 2 has two deadline dates–July 15, 1998, and May 17, 1999. The response to the July 15th date was amazing. Although we have not yet entirely reviewed all of the entries, we already know there were 220 proposals requesting $400 million! (We don't have $400 million yet, but we always have hope.) I will give you a rough proposal profile by content: About 25-30% of the proposals were actually in fundamental research core IT technologies. About another 25-30% were in domain areas combined with core technological research, which is what we want. That would be in GIS, in biology, in physics, in mathematics. There are all sorts of disciplinary applications begging for technologies to help them. Twenty-five percent of our proposals had to do with medicine and health. And surprisingly, almost as many–25%–had to do with the arts and humanities, including proposals for building digital libraries for dance, music, various sorts of ethnographic studies of every sort. It is really wonderful to read these proposals, and we are all just absolutely thrilled by the richness of the offerings.

About three weeks ago, I announced another program to be added onto Digital Libraries Initiative-Phase 2–the Program for Digital Libraries International Collaboration. It has become increasingly evident that doing this locally really makes no sense. So we released a program announcement that we will fund U.S. teams to collaborate with non-U.S. teams. There are no country quotas. This program is, I think, extremely elegant in its simplicity. There are no requirements other than collaboration with a non-U.S. team to be eligible for funding.

Just as a little experiment, about a year ago, I started adding languages to my Digital Libraries Initiative homepage. Now, I have it in nine languages, five alphabets, and three different character systems. The point is that knowledge about whatever topic is encoded in many, many different ways. The non-English portion of the Web is increasing at a much faster rate than the English portion of the Web. In fact, the whole notion of semantics can only make sense when you think of it in terms of information being expressed in different languages, because of their varying capacities for encapsulating and communicating meaning and different types of concepts and perspectives.

This provided some justification for launching the international program. Additional justification came in the form of the success of five NSF-European Union working groups that we fund jointly with our Division of International Programs. The five working groups are in the midst of producing a final report. I believe there is a draft report out available on the Web now, and I think it will also be distributed in the next few weeks. We are quite excited about the international program.

I will wrap this up just by telling you an anecdote. One of the things that I insisted on in the international program was that the currency of comparability of effort was not going to be money. In other words, we don't expect non-U.S. research groups to put in the same monetary amount as U.S research groups. Of course, non-U.S. research funders can put in the same amount, or they can put in more (and I think that many of them will). But they also can put in less. It is the comparability of effort and a really good collaboration that is what we are looking for.

I think the thing that made this idea light up in my head occurred about a year ago, after I joined a telecommunications working group meeting in Russia. On the flight back, I discovered that my seatmate also worked for a nonprofit. I discovered that she and others were working for a Washington, D.C.-based conservation organization that was funding biodiversity studies in Ukraine. When I asked her about the costs of her program, she explained that they gave grants for some Ph.D. researchers from countries in the former Soviet Union to work collecting data at Ukraine sites for a year. The grants, she explained, cover all the researchers’ salaries, all their expenses, all their accommodations, and so forth. I asked the size of the grants, and she responded that they start at $500 a year and go up to $5000 a year! This anecdote illustrates why comparability in currency is something we have to think very carefully. And that is one of the reasons we are looking for international collaborations in different ways.

Previous | Next

Copyright © 2003 NFAIS. All rights reserved. No part of this product or service may be reproduced, stored in a retrieval system or transmitted in any form or by any means, electronic, mechanical, photocopying, recording or otherwise, without prior written consent.

Privacy Policy