JCDL 2009: Digital Libraries: Now here, or Nowhere?

First Keynote
Digital Libraries; Now Here or Nowhere?
Christine Borgman
Professor & Presidential Chair in Information Studies
UCLA

Christine Borgman delivered an opening keynote that explored the past of digital libraries, and inquired into current practices that will keep digital libraries relevant in the future.

As it turns out, the term "digital library" is not frequently used now; its heyday was during the National Science Foundation's Digital Libraries Initiative phases - approx. 1997 through 2004. Since the advent of the Web in 1994, Web architecture and infrastructure have been integral to digital libraries. It's an unfortunate fact of infrastructure that as it becomes more developed and sophisticated, it becomes more invisible (until it breaks), and such is the case with digital libraries.

Computer Science should consider itself an equal partner with collaborators in other disciplines. Problems and collaborative projects have moved beyond CS playing merely a supporting role. Borgman's opinion is that any collaborative project should be equally at home published in a CS journal or a journal of a collaborator discipline, e.g.: Biology. This standard sets the bar very high for collaborative projects, because it requires that the problem under investigation be interesting and worthwhile for all parties. Often, a problem that's interesting to CS is trivial to the collaborator discipline, and vice versa.

Borgman offers the NSF-funded ADEPT project (Alexandria Digital Earth ProtoType) as an example of a past digital library project that contains significant lessons for current and future work (ADEPT and its ramifications were described in a JASIST article 5 or 6 years ago):

• We built it and they did not come, even with significant bribery. Why?
  • A mismatch between faculty course content and contents of ADL (Alexandria Digital Library)
  • A mismatch between ADEPT capabilities and faculty teaching practices
    • The capabilities assumed significant sharing of course content and methods between/among faculty members; in fact, faculty members did not share and did not want to.
  • A lack of university infrastructure
• What did they like?
  • Tools that made their own data useful for teaching
    • Ability to load own data, which was the information they knew best
• Projects like these tend to die because there's no capacity to sustain the project after the initial grant funding is exhausted.

Borgman used a current project, CENS (Center for Embedded Network Sensing), to illustrate the very different needs of various micro-communities that may be involved in a project. For example, engineers who build robots have a very different conception of temperature than do biologists. For a project to be successful, the input and requirements of all these micro-communities need to be accounted for. To that end, a project like CENS must create multiple digital libraries to present data collected by the project to those micro-communities in ways they care about - one size does not fit all. In an attempt to do this, CENS is using ORE (Object Reuse and Exchange) to build aggregation layers on top of various collected data. The problem of appealing to the interests of micro-communities is echoed in Gerhard Fischer's keynote for day 2 of the conference.

An important example of a current project that has some serious implications for future work is the Microsoft Research World Wide Telescope:

• It's open, and it's from Microsoft
• A very sophisticated program, from a software perspective (interesting to CS)
• Sophisticated interface design - now important to sciences too, not just arts and humanities
• A research tool
• A learning tool

Why openness matters:

• Interoperability still trumps all - it answers the problem of making huge cross-disciplinary projects work
• Affects discoverability of related items
• Allows for connections between/among related items, rather than forcing consolidation on related items

So what has changed in digital libraries since the NSF Digital Libraries Initiative?

• The question, "Is it digital, or is it a library?" is mostly passé
• Now here: the scope of work foreseen in the Digital Libraries Initiative definition of "digital library"
• Nowhere: the term "digital library"

Lessons learned so far:

• If we build it, they may not come
• Communities are rarely as homogeneous as they appear
• Community partnerships in project design are essential
• Favor connectedness over consolidation
• Interoperability is still a challenge, but its current flavor is "openness"
• Be open to new opportunities, regardless of terminology