School of Design Image Project: NCSU Libraries

The School of Design Slides Database Project was initiated in 1998 as a prototype for demonstrating the Web's capabilities, both as a means of capturing remote cataloging effort and of effectively searching and displaying large scanned image collections at various resolutions. It has also served as a testbed for the NCSU Libraries' database and Web design skills and our ability to move a project from the design phase into normal production routines.

The first phase of the project consisted of an effort to design the structure for a database capable of handling the Design faculty's need to describe diverse objects in architecture, landscape architecture, graphic and industrial design. At a brainstorming session held in the School of Design on 30 April 1998, an attempt was made to achieve a comprehensive list of possible database attributes. A structured list of these was provided to faculty not present at the session to determine whether additional fields might be necessary in order to describe attributes not currently listed. A second, more prescriptive list, combined attributes which were similar in scope to keep the number of possible fields down, thus reducing complexity of data entry and retrieval. Vocabularies of specific attribute terms, similar to the Getty Institute's Art & Architecture Thesaurus, was planned to control subject terminology, although this has yet to be implemented.

A trial of the database design developed above ended the first phase. Initially, a small number (61) of objects was cataloged to test the design to see where additional attributes needed to be added or current attributes combined, renamed or otherwise rationalized. The NCSU Libraries' Digital Libraries Initiative used ColdFusion to enable browser searches of the test data by keyword, location, agent (creator, designer, architect), and object (building name, sculpture title, etc.). A proof-of-concept search tool, based on this limited database was mounted over summer 1999. A demo Boolean interface for current slide holdings in the Design Library, is described at greater length below.

Original plans called for a more structured test phase to follow the first demo. A test file of approximately 300-400 slides of architectural images was to be cataloged at the School of Design using an ordinary Web browser as the medium for remote data entry. A Web form was designed to enable faculty members or Design students to describe the slide collection at their leisure in the comfort of home or office with resulting data captured in a live database similar to that described above. The form presented images in the left frame while information descriptive of the image was to be entered to the right. Scanned slides from School of Design courseware were presented on the form to demonstrate the scope of data likely to be used during this project. Clicking on one of the slide "catalogs" would bring up an index display (231 x 231 pixels or ca.100 KB) of images to be enlarged by either double clicking on the image itself (medium resolution--ca. 400 x 267 pixels or 250 KB) or on the word "picture" (high resolution--about 1280 x 768 pixels/1.4MB) from the descriptive text.

Current Design Library slide collection

The present Design Library slide collection contains approximately 75,000 slides originally organized using the 1949 American Institute of Architects' (AIA) Filing System for Architectural Plates and Articles with some local amendments to allow for inclusion of non-architectural images and for grouping by time period and place. Until 2000, Bibliographic access to the slides had been provided by means of a card catalog maintained within the Design Library, and included entry by creative agent, title, set title, and subject heading. In the example below, the class number "3Be : T3.1/R82 bs" in the left column may be deconstructed as Baroque Period (3) : Belgium (Be) : Paintings (T3.1) : Rubens (R82) : Bassamento (bs). "S34431" represents the accession number, a unique number assigned to each slide as it is added to the collection.

3Be : T3.1
R82 bs

S34431

Rubens, Peter Paul, 1577-1640, Flemish Baroque painter
Design for the "bassamento" in the Sistine Chapel,
studio of Perino del Vaga (retouched by Rubens c. 1605)
ink and body color on paper. British Museum.
Source: AAOS, May/June 1978

1. Drawing, Flemish--17th century. I. Title.

Many large sets within the collection were accessible only through printed finding aids, as individual slides within the set were not described separately in the catalog and were not classified using AIA.

Until October 1999, there had been no provision for remote or online access to either the images themselves or to metadata describing the slides. The Libraries, recognizing its responsibility for providing better access to this collection, began investigating means of capturing available cataloging information and providing better organization and preservation for the slides themselves. Using available manuscript accession listings, the Digital Libraries Department hired students to enter descriptive information into an MSAccess database over the summer of 1999. Relational files were added to control the vocabulary for agent names, medium/form of art, and place. This simple accession listing formed the basis for the current much enlarged database, enabling online keyword access to the slides by title, place, creative agent, medium, source of image, set title and number, classification and accession number for the first time. A demo Web Boolean search form was designed by staff in the Libraries' Systems Department, using ColdFusion to pass against the raw MSAccess database, and was demonstrated to Design faculty on 22 October 1999. While this demo lacked digital images and the database demanded quite a bit of cleanup, it still demonstrated the power of Boolean searching across multiple fields in a surrogate file. The database framework should also accomodate the later addition of structured vocabulary terms from the Art & Architecture Thesaurus, Library of Congress Subject Headings or Thesaurus for Graphic Materials, and any other list as demand dictates.

By late 1999, the Design Slides database consisted of an inventory/citation table and several relational tables. These contained the following fields:

Citations (Main table)	Agents	Places	Medium
Accession # TimePeriod- numeric code Placecode- 2 letter code for place AIAClass- form/genre-based class Title- often descriptive text SetTitle and number Date- date of creation of piece depicted View- aspect of work Vendor- source of purchased slides ImageSource- for copy work, privately generated slides	Accession # Agent- name of artist, designer, firm Vocation- cabinetmaker, architectural firm, etc. Nationality Type- personal, family or corporate name Source- authority file used, usually ULAN or LC	Accession # Place- building, community, state, country, body of water	Accession # Medium- form of object on image

In July 1999, the Libraries hired Rachel Kuhn, as the first in the initial crop of five Libraries' Fellows. Rachel had had experience within a larger slide collection at Wake Forest University, and worked with Cataloging, DLI and the Design Library towards reorganizing and providing better remote access to the Design slide collection. Working with Design faculty and library staff, Rachel developed a strategic plan to deal with the slides collection and began working to add sets not classified in AIA to the accession database. She was assisted by Hayley Charney, a Ph.D. student, who did much of the cleanup work on the original student-created files, and verified information from the accession listing against the old card catalog and the actual slides themselves. The slow task of having slides digitized in priority order, at first through an outside vendor, but later in-house, was begun in Winter/Spring of 2000. The Libraries' Systems Department also became more actively involved in the project at about this time when the team of Shirley Rodgers, Rob Main and Troy Simpson ported the data over to Oracle, created a data entry front end and began working on an improved user interface. By 2001, a second Libraries' Fellow, Kim Duckett, was hired to work on the slides project and, together with Rachel, helped catalog thousands of slides and images into the new database. Later Jamie Vermillion was hired by the Cataloging Dept. to join this effort on a half-time basis. He was followed by Hayley Kyle in 2004.

The present Design Library Image Collection search engine is built around the original, now greatly cleaned up and enlarged, database. It includes digital images in both thumbnail and larger sizes plus a number of new features specifically requested by end users, such as "Portfolio builder". This feature allows the user to add images to a virtual shopping cart and to build collections for research purposes, classroom slide shows, and so on. A circulation module, utilizing barcodes added to the slides, has also been added to allow for the circulation of slides to Design Library patrons. Three years after the first data entry began on this project, the new resource discovery tool was unveiled to the public as yet another NCSU Libraries database.

Slide classification project

Libraries have used standards to organize and describe their collections of books, periodicals, media materials, and now computer files, for well over one hundred years. This is in large part due to the nature of their contents, which generally cover the universe of printed knowledge and tend to describe that universe textually. Much of this standardization can be traced to the strong influence of the library world's professional organizations, which seek economies of scale through cooperative effort. The leadership of the large national libraries, such as the Library of Congress and British Library, has also exerted a powerful homogenizing force.

Slide collections, like book collections, also occur in diverse institutions, yet here the success of standardization efforts in organizing and retrieving materials has been disappointing. Even the most successful effort towards this end, the Getty Information Institute, creators of the Art and Architecture Thesaurus and Union List of Artist Names, has been forced to close down. There are many reasons one can give for our inability to provide uniform retrieval tools for slides. Foremost, slide collections, unlike those for books, are often comprised of non-published materials created as copy work from the pages of books and magazines, or captured by or for the private use of individual teachers or researchers. There can be no economy of scale when each institution's collection consists largely of unique images; images for which copyright is somewhat nebulous. Copyright is often a disincentive in the sharing of bibliographic information for images, both where the rights holder is not known, as is the case for copywork and private collections, and where they are known, as with purchased slides and sets. Libraries and museums don't want to confront copyright holders by publicizing their individual holdings broadly, so largely keep surrogate information, such as image descriptions, to themselves. This is doubly true for those who also provide their patrons with digitized surrogate images as part of the retrieval mechanism.

There are other factors weighing against standardized retrieval tools for slides. Slides do not describe themselves in the way that books do. There is often no text on the mount to describe who built, designed or painted the image contained therein; where it is located, or what its actual size is. Often it is only the creator of the slide, who may have been a friend of a former colleague of the donor, who knows what is depicted. Donors and user populations for slide collections also each have their own perspectives on how the drawers should be optimally organized: by period, by artist, by structure, by form, by place. A survey of slide collections at other institutions reveals the wealth in variations of organizational models, even amongst those claiming to follow a particular organization standard. And of course, once a model is in place, it is very expensive to reorganize collections of tens, or hundreds, of thousands of slides.

The NCSU Libraries has investigated a number of general slide classification schemes over the past year, in the hope that we might be able to adopt a model here without the additional effort of having to invent our own. Unfortunately, there is not even a standard way of implementing the schemes that do exist. Each museum, library or art school has adapted the features of the classification that fit their own needs. This is no doubt due to the fact that none of the slides-only classification schemes exist in a formal published format such as exist for the Universal, Dewey Decimal or Library of Congress classifications. Rather, systems seem to migrate from one site to another through word-of-mouth, staff manuals and handouts, or articles published in the likes of the VRA Bulletin.

In reorganizing the Design Library slide collection, we have devised a slide classification system based on a combination of features from the AAT and several implementations of the Fogg Museum Classification. One of the shortcomings of the present AIA-based classification is that its hierarchy is based first on period and place. This disregards the nature of different art and design genres, which often have periods specific to that form and which may be more conducive to non-standard hierarchical subdivision. In dividing first by format, we can also enable more generalized classification patterns for areas of low collection activity such as cartography and the decorative arts, while areas such as architecture are provided a more detailed organizational framework. Comments on this model may be directed to Charley Pennell, Head of Cataloging. Use of this new classification scheme for organization of slides within the Design Library began late in 2000 and continues to the present, with new slides being immediately classed into the system while older slides are handled in demand priority.

QuickSearch:	FAQ\|Site Index

The School of Design Slides Database Project

Current Design Library slide collection

Slide classification project

Links