Library of Congress

Program for Cooperative Cataloging

The Library of Congress > Cataloging, Acquisitions > PCC > PCC Standing Committee on Automation > PCC Standing Committee on Automation: Task Group on OCLC Batchloading

Final Report

February 28, 1999 (rev. March 17, 1999)

The OCLC Batchloading Task Group (Carol Hixson (UCLA), Chair, Mechael Gago (Indiana University), John Levy (Library of Congress), Gary Strawn (Northwestern University), and Jay Weitz/Robert Bremer (OCLC)),of the PCC Standing Committee on Automation was charged in 1998 to advise OCLC on issues surrounding the batchloading of bibliographic and authority records, especially as they related to PCC libraries' contributions. The Group met at ALA Midwinter in New Orleans and ALA Annual in Washington, D.C. and carried out other work via email and phone calls. After considerable discussion, the Group made a series of recommendations, to which OCLC responded, as follows:

  1. TG: OCLC needs to implement Encoding Level 4 fully to facilitate the batchload process for Program records.
    Background: Encoding Level 4 is the USMARC code (Leader/Byte 17) for Core records. OCLC is not implementing value 4 at this time for online input. Instead, OCLC has implemented field 039 for Core-level standard identification. Field 039 is not output as 039 and is a convention for online input only. Tapeloaders to OCLC use USMARC practice when sending files to OCLC. In OCLC-MARC output (via OCLC-MARC Subscription tapes, exported records, EDX files, etc.), OCLC converts field 039 core to Encoding level 4. With the conversion of Encoding Level 4 to an 039 textual field, the ability to carry out batchloaded matching and merging of PCC records against lower-level or non- Program records already in the OCLC database is not feasible.

    OCLC response: this is on hold until completion of Year 2000 and Authority Control Rewrite projects that are underway at OCLC.

  2. TG: OCLC needs to provide more equitable credits for batchloaded records compared to records directly input.
    Background: The current difference in credits between batchloaded and directly- input records is substantial and serves as a deterrent to using the batchload process for submitting Program records.

    OCLC response: while the credit is small, it is growing. If libraries use CatME, they will receive the identical credit as for direct input. OCLC now provides CatME free of charge to member libraries. OCLC ties credits to timely availability of records and timeliness is currently a problem with batchloading. With CatME, libraries are able to upload records from their local databases and send them as a batch to OCLC and receive credits that are identical to those for records that are directly input.

  3. TG: OCLC needs to process batch files within 24 hours of receipt.
    Background: The primary goal of the Program for Cooperative Cataloging has been, from the outset, to increase the timely availability of bibliographic and authority records, as well as to reduce duplication of effort. With the delay that is experienced with batchloading, it is possible for another Program participant to submit, via direct input, a record for the same item while the batchload file is waiting to be processed.

    OCLC response: much of the batchloading process requires manual intervention and 24 hour turnaround is not practical. OCLC is working on automating the process more as part of their Batchload Redesign Project and is also looking at other ways to increase timely availability of records, including a Z39.50 original cataloging capability.

  4. TG: if a batchloaded PCC record matches an existing record in the OCLC database, it is recommended that OCLC either attach the batchloading library's holding symbol to the existing record and notify the library of the match so that it can decide whether or not to carry out a manual upgrade, or merge the two records, retaining all national-level data in the merged record.
    Background: Currently, if a library sends a Program record as part of a batchload file and the Program record matches against an existing record in the OCLC database, the batchloaded record is not accepted (except for holdings information). If the batchloading library still wants to upgrade the existing record to Program standards, that library must lock and replace the existing record manually, rekeying the information and attaching the library's holding symbol to the existing record.

    OCLC response: OCLC is working on this. It is related to the full implementation of the Encoding Level 4. OCLC suggested that the Task Group could help to identify the fields to be retained in such a merged record.

  5. TG: OCLC needs to allow for a batchload lock and replace mechanism, matching on OCLC number and checking the 005 to determine the later version of the record. This is of particular interest to the Library of Congress regarding CONSER records.

    Background: Program libraries would like to be able to download non-Program records into their local system, edit them there, and upload them as locked PCC records to OCLC as part of a batchload lock and replace. Currently, lock and replace works only with direct input.

    OCLC response: it is hoped that the Batchload Redesign Project will allow for improvement in this area. OCLC would like the Task Group's help in adjusting the CONSER field selection criteria.

  6. TG: OCLC needs to accept a mixture of PCC and non-PCC records in the same batch of records.
    Background: Most Program libraries do a mixture of PCC and non-PCC cataloging. For those libraries who prefer to do cataloging in their local system and batchload those records to OCLC, the inability to send a mixed file to OCLC requires either separate workloads or local programming to generate separate files to send to OCLC.

    OCLC response: This is on the list of issues to be dealt with in the Batchload Redesign Project.

Following the OCLC responses to the recommendations, the Task Group identified three possible areas for continued work. Chair Carol Hixson discussed these with the PCC Standing Committee on Automation at the ALA Midwinter meeting in Philadelphia on January 31, 1999. Regarding the possibility of working with OCLC to identify fields to be retained in a merge process, it was decided to defer work on this until OCLC finishes its Batchload Redesign Project. Regarding the need to work with OCLC to adjust CONSER field selection criteria for a batchload lock and replace, it was decided to defer work on this until the Library of Congress has implemented its new ILS and the specific needs and limitations of its new system are known. Regarding the testing of the newest version of CatME to determine if it serves as an effective interim solution to the needs identified by the Task Group, David Whitehair noted that he had received three test files and had made adjustments to the program based on those files. It was decided that the Chair of the Standing Committee on Automation, Karen Calhoun, would post a message to the PCCList and the PCC Home Page to alert Program participants of the real time uploading feature of CatME, asking them to test it in their libraries, and inviting comments to be posted to PCCAUTO (the Committee's listserv). The announcement will include a link to the download site for CatME software.

The members of the PCC Standing Committee on Automation thanked the Task Group for their work. They felt the TG had completed its charge to the extent possible at this time. Therefore they requested the TG's final report, which they will forward it to the PCC Policy Committee with a request to discharge the group. The SCA also recommended that a new TG be appointed once OCLC is ready to resume work on its batchload redesign.

Back to Top