Draft report on a meeting to explore possibilities for developing a shared VR cataloguing service hosted by NYU with support from the DLF on 22 January 2001

Present: Lee Chamberlin (NYU), Murtha Baca (Getty), John Weise (Michigan), Linda Barnhart (UCSD), Greg Zick (Washington), Sherman Clarke (NYU), Michael Ester (Luna Imaging), Heike Kordish (NYPL), Jane Murray (NYPL), Susan Williams (Yale), Chris Sundt (Oregon), Elisa Lanzi (Smith), Ann Whiteside (Harvard), Daniel Greenstein (DLF)

This report summarizes consensus that emerged at the meeting about the design, use, and benefits of a visual resources cataloguing service, and the next steps that should be taken in the development of that service.

For the benefit of participants, the report concludes with comments that distinguish this initiative from others in which the DLF is involved, notably the Academic Image Cooperative and the Open Archive Initiative metadata harvesting testbed.

1. Aim of VR cataloguing service

(Note, the term "service" is currently under discussion and may well be replaced by "workstation" or "tool").

The service will support and promote the shared development of high-quality, consistent, reliable, and professionally developed descriptions of visual resources.

2. Service development and functions

• The service will consist initially of VR descriptions supplied by a small number of institutions.
• Initial contributions will be solicited in order to ensure the broadest possible thematic coverage and the minimum degree of overlap.
• Contributed descriptions will be stored in their native formats and presented either in their native formats or in some unifying view to be developed.
• The service will include a range of tools with which users can clean, enhance, amend, and "improve" existing descriptions according to agreed guidelines. Tools will also allow users to create and add new descriptions, again, according to approved guidelines. A list of editing tools is supplied below.
  • Tools allowing some users and/or service administrators to cluster descriptions that refer to a single work. Some research will need to be conducted into how and to what extent that clustering can be automated and how clustered descriptions can be displayed effectively.
  • Tools enabling users to search and browse descriptions and clusters descriptions that refer to the same work
  • Tools enabling users to edit, amend, and enhance existing records including:
    • pick lists that assist in the selection and application of controlled vocabularies and vocabulary key words;
    • mechanisms for including/uploading thumbnail images or links to web-accessible images where images are considered to be key descriptive elements.
  • Tools enabling some users and/or service administrators to upload collections of image descriptions in their native formats.
  • Tools that allow users to output records for local use, e.g. as printed labels, in standard file formats for inclusion in a local catalogue database, as links to authoritative records in the catalogue.
  • Tools allowing users to suggest key terms and to submit these for review by some approved editorial board.
  • Tools allowing users to group, interpret, supply, and "publish" as part of the service additional contextualizing information for individual images and image sets.

3. Levels of use

The service will permit different levels of use including the following

• Level 1.Users will be able to search, browse, and export records (e.g. for printing labels, populating local catalogue databases, or supplying links to authoritative records in the catalogue).
• Level 2. Level 1 functions plus ability to add descriptions. Added descriptions may provide supplemental material for works for which descriptions already exist. Alternatively they may supply records for new works. Level 2 users will also be able to submit candidate authority terms for existing or newly created descriptions.
• Level 3. Level 2 functions plus ability to approve and implement authority terms. In this respect, at least some level 3 users are likely to serve as members of specialist editorial panels or review bodies and take responsibility for reviewing and approving key terms submitted for review by level 2 users.
• Level 4. System administrator.

4. Audience(s) and benefits

The catalogue service will have a number of direct benefits for VR cataloguers and slide and art librarians. For members of this community it will:

• reduce redundant cataloguing effort;
• encourage, support, and provide examples of good cataloguing practice and in this respect provide an important professional training and leadership function;
• ensure convergence of and leverage complementary but fragmented streams of activity within the VR community e.g. on controlled vocabularies, agreed metadata formats and application guidelines, etc.

The service may also have a number of indirect benefits for other communities as follows:

• As cataloguing effort is shared, VR collections will become more accessible and as such will do more to support and enhance teaching and research. In these respects, the service will benefit students, teachers, and scholars.
• Scholars, students, and teachers may also benefit through some direct access to the cataloguing service or to some middleware applications developed by VR and library professionals. There are a number of research questions here about what access might be provided to scholars, students, and teachers and/or what middleware applications might be developed.
• The catalogue service, by providing a common set of high-quality descriptions, might encourage greater collaboration between libraries, special collections, archives, art libraries, museums, and galleries that share a common interest in the visual resource. How a service initially targeted at the community represented by the Visual Resources Association might benefit this broader group is at least initially a research question.

5. Scenarios and incentives

Description contributors. Institutions with online VR catalogues will be more likely to contribute records to the service if the service supplies them with tools capable of enhancing records in those catalogues.

Contributions from trusted "level 3" users (those specialists able to add or approve authority controls) may be sought in a variety of ways:

• By leveraging and supplying good cataloguing tools to existing catalogue initiatives whether they are taking place within a single institution (e.g. the Huntington collection) or in a consortium (e.g. AIC).
• By integrating this kind of activity into fellowship, internships and other professional development and professional training activities.

Service developers. Previous efforts in this area have been limited by their heavy reliance on voluntary effort. Accordingly, any initiative that may be undertaken to fulfill the goals set out in this document should seek appropriate support and develop from the outset with organizational and financial sustainability as a pre-eminent goal.

6. Technological issues

The service will need to investigate use of different approaches for updating and searching respectively, possibly putting updating on an incremental as opposed to a dynamic basis

Design of an effective user interface will be challenging given that the service will need somehow to present users (cataloguers) with multiple descriptions, authorities and vocabulary choice, etc

Architecturally, the service should be developed in a manner that will cost effectively support the evolution and extension of its functions

7. Next steps

Steps to be completed by end March 2001

• Host focus groups at upcoming VRA and ARLIS meetings to get input from VR cataloguers about the tools they would wish to see in a cataloguer's workstation.
• Investigate architectures and functionality of existing cataloguers workstations and tools as developed for bibliographic materials e.g. RLIN, OCLC, LC.
• Develop an initial implementation of a common record view; that is, the view with which all descriptions will be available from the service irrespective of native format.
• Develop prioritized list of controlled vocabularies that need to be included in the service.

Preferably steps listed under 2 should completed by end April 2001

• Document costs and requirements of an initial prototype.
• Present initial thinking about the service to members of communities that may have an interest in such a service e.g. museums and galleries, archives and special collections, etc. Will serve important community building and design functions.
• Assemble initial contributions of VR descriptions.
• Develop and review prototype (timing contingent upon 2 above).

8. Relationship of VR cataloguing service initiative and other potentially complementary DLF initiatives

1. Open Archives Initiative (OAI)

   The OAI is an international initiative that has developed a metadata harvesting protocol. The DLF supports the OAI but is not the OAI. Information about the OAI is available from www.openarchives.org

   Using the OAI metadata harvesting protocol it may be possible to assemble VR descriptions stored and managed in a single place. In this respect, metadata harvesting might provide one means for a cataloguing service to assemble raw VR descriptions.

   The metadata harvesting protocol is relatively untried; its application is highly experimental.

   In the meantime, the real effort for the VR cataloguing service is not assembling VR descriptions but developing the tools for editing, enhancing, and disseminating descriptions.

   Accordingly, although the OAI harvesting protocol might contribute a single, relatively small component to a VR cataloguing service, its experimental nature suggests that any prototype work be conducted on the basis of physical assembly of descriptions.

   This should not preclude work on experimental or prototype harvesting services that focus on visual resources. Indeed the DLF is actively hoping to encourage the development of such services.

2. Academic Image Cooperative (AIC)

   With funding from The Andrew W. Mellon Foundation and the support of the College Art Association, the AIC was initiated in January 1999 as a planning process to develop a scaleable database of curriculum-based digital images for survey courses in the history of art. The planning process was completed formally in August 2000 and resulted in a prototype database and image collection. It also developed technical, organizational, and policy frameworks that have the potential for sustaining a more ambitious online service; one capable of identifying, developing, and disseminating a far large number of curriculum-based and scholarly image collections.

   Since the completion of this planning process, ongoing discussions between the DLF and The Andrew W. Mellon Foundation, have focused on how the AIC's image collection - and the learning derived in its development - might contribute to a broader initiative under consideration by the Foundation and provisionally named ArtSTOR. Briefly, ArtSTOR is looking to do with visual resource materials something analagous to what JSTOR did for periodical literature in the humanities and social sciences: create an essential digital library that responds to widespread scholarly needs.

   The result is a collaboration whereby the DLF is helping the foundation to develop circumscribed, strategically identified image collections that respond to widespread teaching and other specialist scholarly needs. It is envisaged that these collections, including the one developed by the AIC, will be incorporated into the evolving ArtSTOR service, which eventually will be managed as a project of the Foundation or an organization that they designate.

   For a programmatic and preliminary description of the ArtSTOR initiative, please consult the 1999 President's Report available from the Mellon Foundation website at http://www.mellon.org/President annual report 99.pdf. These pages document the AIC's development as a DLF initiative.

3. VRA implementation guidelines

   The DLF wishes to support the VRA in its development of implementation guidelines. Obviously such guidelines would make an important contribution to any VR cataloguing service. Work on the service is viable in advance of any such guidelines being produced.

   return to top >>