History of the Project
An OPAC was installed at Princeton University in the Firestone Libraries in 1981. Approximately six million pre-1981 catalog cards were not converted to the OPAC, but were still very much used. These catalog cards, stored in 5,107 drawers, represent 1.75 million items owned by the library.
Several years ago, the Princeton University Library and Princeton's Computer and Information Technology Department (CIT) began to search for ways to preserve and convert the pre-1981 public card catalog. The goal for the project was to automate the card catalog in order to provide patrons a more easily accessible information resource.
A number of vendors were solicited for information regarding many conventional retrospective conversion practices. The responses received indicated that it would cost millions of dollars and take years to complete the conversion project. In addition, Princeton had not updated its shelflist and therefore needed to ship the public catalog cards to the vendor. This would have left the library without a public card catalog for the duration of the project. Digitization of the card catalog kept entering the picture so university officials pursued an intensive evaluation of imaging options and systems developers.
Proceeding with the Project
As with any project that incorporates the use of new concepts and new technology, the library needed to show monetary as well as operational benefits before undertaking such a large-scale project. Significant factors in the decision to proceed with imaging were:
1. Labor costs for handling the catalog cards had increased and would continue to do so.
2. A study conducted on a randomly drawn sample of 1,500 cards indicated an 11.13 percent deterioration rate of the cards, with a margin of error of plus or minus 3 percent when projected over the entire card catalog. The cards evaluated for this study included handwritten, typed, Library of Congress, and OCLC cards. One of the problems encountered was the old card stock, but there were also problems with newer cards that were used frequently
3. The bids received from the hardware vendors showed that storage and computing costs were rapidly decreasing, reducing the storage costs for the images considerably
4. Optical character recognition (OCR) technology was continuing to improve.
5. Retrospective conversion was too costly and time-consuming.
6. A strong commitment was received from VTLS Inc. that the project would be satisfactorily completed under the budget guidelines.
The benefits of this project were that the information on the cards would now be preserved and easily accessible by Princeton's patrons. Both benefits coincided with the library's mission. These benefits alone gave the library administration the momentum to proceed with the project. An added benefit of the project was the last phase. In this phase the library would use optical character recognition technology to convert each digitized image to fully tagged and indexed records of text in a standard MARC format.
The project plan outline was as follows:
1. Clean the card catalog. 2. Index the card catalog. 3. Scan the cards. 4. Quality control of the scanned cards. 5. Rescan rejected cards. 6. Quality control of the rescanned cards. 7. Implement the VTLS Image-Browser. 8. Train staff and faculty to use the VTLS ImageBrowser. 9. Convert the scanned images into a MARC format using OCR technology
Selecting a Vendor - VTLS
When Princeton submitted proposals to vendors, we already knew our needs. We were looking for a vendor that could manage the entire project - one that could supply a turnkey solution. The vendor chosen needed to have expertise in imaging hardware and software and to have or be able to develop the unique software applications required for this project. …