2nd Annual joint conference on digital libraries

Citation

Seadle, M. (2002), "2nd Annual joint conference on digital libraries", Library Hi Tech News, Vol. 19 No. 10. https://doi.org/10.1108/lhtn.2002.23919jac.003

2nd Annual joint conference on digital libraries

Michael Seadle

Overview

The Joint Conference on Digital Libraries (JCDL) is a combined effort of the ACM (Association For Computing Machines) and the IEEE (Institute of Electrical and Electronic Engineers). These two sponsors each bring somewhat different traditions, but they seem almost to come from the same tribe compared to the much broader range of research cultures represented by the holders of Digital Library Initiative (Phase 2), National Science Digital Library, and Institute of Museum and Library Services (IMLS) grants. From an anthropologist's viewpoint, the interactions (and avoidances) are as interesting as the research itself.

It is a strength of the conference that there are in fact no fixed boundaries between groups and session topics. As a library person with strong computing interests, I found myself shuttling between upstairs sessions that talked about human-computer interaction issues such as building interfaces, and downstairs papers and panels on metadata and preservation. Some attendees stayed in one room, but peripatetic behavior was far from unique.

The long breaks between sessions and the shortage of tables in the refreshment room facilitated additional interactions. Those associated with giving away grant and foundation money had constant companions, and at times took on the weary look of those longing for non-business chat with old friends. But they were available in a very deliberate, very professional way that emphasized the openness of their agencies to fresh faces and new ideas. Newcomers had an excellent opportunity to question both agency officials and the people who serve on review panels. Many novices may overlook the fact that the latter matter as much as the former.

About the sessions

The conference had too many excellent sessions to report on them in full. I will mention only three.

OAI for EAD

Chris Prom and Tom Habing from the University of Illinois at Urbana/Champaign (UIUC) talked about the use of Open Archives Initiative (OAI) protocols with Encoded Archival Description (EAD). UIUC has been a leader in EAD use with OAI metadata harvesting.

They argued that OAI is a good tool for cultural heritage materials, and is more established than METS (Metadata Encoding and Transmission Standard). OAI interoperability is based on a harvesting process, rather than distributed searching. This allows the harvester to generate multiple OAI records for a single EAD file, depending on the depth and complexity of the tagging.

One problem with these multiple records is maintaining context. Although XPointers can help to do this, browsers still lack XPointer support. For EAD records with well developed collection-level information, a single OAI record may be preferable. Encoding practices vary widely, but as more institutions create OAI-based collections, practices may become more uniform.

Digital preservation

Joyce Ray from IMLS moderated a panel on digital preservation. The speakers were Vicky Reich, Director of the LOCKSS Program at Stanford University Libraries, Robin Dale, Program Officer for Member Initiatives at the Research Libraries Group (RLG), William Underwood, Principal Research Scientist with the Information Technology and Telecommunications Laboratory at Georgia Tech Research Institute, and Reagan Moore, Associate Director for Data Intensive Computing at San Diego Supercomputer Center.

Joyce Ray opened the discussion by noting two significant recent developments in digital preservation:

1. 1.

   it is now recognized as a legitimate field for research in computer science; and

2. 2.

   we are also recognizing that there is not a single solution to digital preservation – organizational and economic issues are as important as the availability of technology and tools to support preservation.

The discussion revolved around three themes: preservation technology, metadata, and authenticity. The LOCKSS (Lots of Copies Keeps Stuff Safe) project offers a digital means of reproducing one of the key preservation techniques of the analog world, where so many libraries have copies of important works that local disasters do not affect availability. The LOCKSS servers share data and do open peer-to-peer checks to ensure that content remains unaltered and undamaged. LOCKSS does not guarantee that future systems will be able to use the data unaltered, but maintaining the copies on active disk drives with constant integrity checking does meet one of the key concerns of preservation professionals.

Preservation requires more than maintaining data integrity. Open Archives Information System (OAIS) provides an intellectual framework for the developments of standards, and can provide an architectural basis for software design. It is not, however, currently a software package that can immediately be applied. METS is an outgrowth of OAIS. It is designed to collect the kind of information that a repository needs to know about digital objects. Its purpose is less to enable their discovery during a search process, though its descriptive section certainly fosters that, than to answer basic questions about what the object is, how it was created, its provenance, and any intellectual property restrictions.

Authenticity depends less on software and metadata, though both are important, than on careful documentation and reliable procedures. William Underwood used the hard disks from the Clinton White House as an example. They were taken on the last day of his administration by subpoena and given to the National Archives and Records Administration (NARA) for safe keeping. Fakes are a problem. They turn up with regularity in the analog world, where they are relatively hard to create and where we have tools that effectively uncover them. Fakes are much easier to make in digital objects, unless their origin and handling can be guaranteed.

Some of the discussion centered on the necessity or desirability of maintaining the original "look and feel" of a digital object. Reagan Moore compared maintaining objects in their original form and format to reading sixteenth-century books by candlelight. William Underwood noted that archivists might feel that maintaining the original look and feel is sometimes necessary in order to preserve the authenticity and credibility of a document as evidence. Joyce Ray summarized by noting again that the field of digital preservation has progressed – different preservation decisions will be called for depending on the content, the reasons for preservation and intended use.

Daniel Greenstein's Keynote

Daniel Greenstein is the Director of the California Digital Library and was until recently the head of the Digital Library Federation. He spoke not about particular digital library projects or issues, but about how people think about digital enterprises. He began with his training as an historian and talked about how libraries have had several distinct periods of digital involvement, starting with MARC record creation and online catalogs and running through a passion for the Web and a desire to digitize "stuff" of every sort.

While for some, content seems paramount, all too many grant applicants claim in their proposals that they have the greatest collection that must be put online at once. For others, infrastructure matters most, but they let their advocacy for certain tools acquire a religious fervor that robs them of perspective. The measurement of uses and users now seems to be replacing (or at least moderating) the unbridled passion just to get stuff online, which is probably a positive sign.

Daniel Greenstein sees a number of key issues for the coming years, including fostering more applied research, establishing a new means of scholarly production, and building the digital equivalent of the interstate highway system.

Conclusion

JCDL is not for everyone. I have colleagues who complain that papers are too technical for them to understand, and others who complain that the acceptance standards have lost their rigor in an attempt to include papers from non-computer science types from soft areas like libraries. Probably both complaints are true, and their balance offers some measure of the success the program committee has had in making the conference a significant event for the digital library community as a whole.

Michael Seadle (seadle@msu.edu) is Head of the Digital and Multimedia Center at Michigan State University Libraries, East Lansing, Michigan, USA.