Cataloguing by Crowd; A proposal for the development of a community cataloguing tool to capture subject information for images

Susan Chun, The Metropolitan Museum of Art, USA
Michael Jenkins, The Metropolitan Museum of Art, USA

Professional Forum: Cataloguing by Crowd

Art museums seeking to publish their collections online generally support user retrieval with browse and search tools served by existing cataloguing data: maker information, object dates, medium, technique, classification, dimensions. Although this catalogue data — often from a museum's collections management system — can be extensive, it rarely contains information about the subject of a work of art: the people, places, ideas, emotions, and events depicted or represented. In some cases, subject information (a natural point of access for both specialist and non-specialist users), is available only through searching of unstructured descriptive text, returning unwieldy or inaccurate results. Often, subject information simply does not exist.

The co-conveners will provide and briefly summarize background materials. We will present a history of past projects aimed at capturing subject or iconographic information (the all-volunteer cataloguers at the Fine Arts Museums of San Francisco; the professional subject cataloguing staffs at European institutions such as the Tate and the RMN; experimental web-based projects such as ESP game) and cite other online community cataloguing initiatives (Wikipedia, Project Gutenberg, the many communities cataloguing song lyrics). We will review studies of the ways in which users of online collections search images and discuss the considerations raised for publishers of content by such tools as Google's Image Search. Finally, we will present a proposal for the development of an open source tool for collecting subject information for works of art in our collections and for related tools to validate and process the collected subject terms. (It is possible that we will have a working prototype available for demonstration and discussion in April 2005.)

The remainder (in fact the bulk) of the session will be devoted to a discussion with audience members about how such a tool might be developed to serve the museum community at large. We will consider the various communities implicated in such a project (developers; content holders, including museums and their curatorial staffs; volunteer and paid cataloguers; end users), debate the value of open source development; and discuss the ways in which data might (or might not) be evaluated, validated, processed, and applied. The value of existing classification systems such as ICONCLASS, and their role—if any—in enhancing community-sourced data will be examined. Participants are encouraged to consider and critique the MMA proposal and to present their own ideas for further collaboration.