eBooks and Ex Libris Voyager

eBooks are an easy way for libraries to add valuable content at (usually) a discounted price from print monographs. They are particularly useful for areas of the collection, like computer science, that seem to be continually updated. My library has had a very strong eBook collection, that due to a mis-assigned professional, had been cataloged and uploaded (by hand!) incorrectly  for entry into our OPAC.

For those unfamiliar with the Voyager ILS the catalog record consists of three pieces. The BIB record that includes all the bibliographic data in MARC format, the HLDG record that displays local location and call number data, and finally the ITEM record that includes location data as well as an individual items barcode.

Library administration felt that the easiest way for collection development to keep up with the changes in academic disciplines as well as patron demand was to set up a Patron Driven Access (PDA)  plan for eBooks from the largest vendor that we subscribe too [~70000 eBooks.]

With PDA plans a library will purchase limited access to a title that become a permanent purchase once the title is accessed a preset number of times. 

This created a difficulties in that we would need the ability to add or delete large number of records to our catalog. First off the eBooks already in our catalog had been created with all three levels of record present. Voyager does not allow bulk deletion of records with ITEM records. We decided that all eBooks would need displayed in a similar fashion so that all ITEM records attached to eBook holdings, no matter the vendor, would need to be deleted. This was done by our in-house developer/server wizard who simply altered CSS on the OPAC display so that any record without an ITEM record would be displayed in a similar fashion.

Within the record users will find a link created from a URL within the MARC record 856 field directing them to the desired content. 

To delete 120,000 ITEM records a list was compiled using Georgia State's Voyager Reporting System (VRS) of all the ITEM records that had been given DUMMY as a barcode. These where then entered into the Voyager Cataloging module using a MacroExpress job that deleted the ITEM and HLDG records.

[Since this initial job took place a script was developed that does the same job on the server, which is much faster than MacroExpress.]

Records where then downloaded from the PDA vendor's website. Massaged using MARCedit, so that they fit library standards and had proper proxy credentials, and uploaded to the OPAC via the bulk import method. To ease processing of these titles, I used the vendor identification present at the end of the URL as the MARC 035 [other identification number] in place 

Since we did this initial load of PDA books, budgetary concerns led the library administration to decide that a cap on the price of PDA books would be set. This made it necessary to delete all the updated PDA records as the vendor could not provide a detailed list of changed titles. This was done by extraction of the vendor ids from the 035 field and then matching these numbers with the BIB ids created by the previous bulk importation.These ids where then imported into Gary Strawn's Cataloger's Toolkit to delete the affiliated BIBs. The toolkit allows manipulation of Voyager records on a greater scale than the 10000 limit suggested for bulk changes using the Voyager Server Bulk Utility. 

Once all BIBs where deleted, an updated list was downloaded from the vendor, manipulated via MARCedit, split into manageable pieces and uploaded to the server. 

I negotiated with the vendor to manipulate future updates so that the 856 URL is pre-populated and the vendor id is present in the 035 field. The small month updates and deletions can now be managed via the Voyager Server Update due to field matching rules set up in the upload rules.