Authority Control Advisory Committee

Authority Control Advisory Committee Minutes

Date: 2002-04-29

Present: Ellen Ellickson, Xinkai Kong, Beatrice Luh, Cécile Mandour, Earl Roy, Rick Sarcia, Manon Théroux (chair and recorder).

Absent: Audrey Novak, Dajin Sun.

1. MARS Project

1xx Flip. The Systems Office has completed the "1xx flip". This fix allowed the 1xx transactions that were inadvertantly loaded after the 2xx fields to be restored to their proper place in the MARC record.

Gap Extract. The Systems Office completed the "gap extract" on April 8th and we received the correction transactions from MARS on April 25th. The transactions will be loaded before the migration. All bib records with a 928 field marked "actape" were included in the extract except for 1) Pinyin conversion records and 2) those CJK recon records that were loaded before we started sending our CJK recon records through MARS. It would have been a waste of money to extract these records because we have ordered a snapshot of our CJK records from RLG. The records in this snapshot (scheduled for April 30) have both romanized and vernacular fields and will be loaded into Voyager as part of our migration. The snapshot records would have overlaid the records sent through gap processing and wiped out any changes. The snapshot records will have to be extracted from Voyager at a later date and sent through MARS. One potentially confusing thing: the Pinyin and CJK recon records that were excluded from the gap extract still had their 928 field changed from "actape" to "ac[date]".

New Profile for Post-Gap Processing. The "gap" records are the last batch of Yale records that will have transaction records generated as a result of processing. Post-gap processing will use full record replacement. The Library will be setting up a new MARS profile for ongoing processing that will be very similar to the profile used for our recon records, which also used the full record replacement approach. One definite advantage to using full record replacement is that we can handle split headings better. The transaction loader could only handle splits where one heading split into two; splits of more than two have to corrected manually. Another possible advantage is being able to use MARS to correct filing indicators in 245 fields. We will need to QA the first batch of records processed using the new profile.

2. Voyager Implementation

Freeze on Auth Record Deletion by Staff / Purge of Deleted Auth Records from ORBIS . On April 15th, staff were instructed to stop deleting authority records in ORBIS until further notice. This restriction was triggered by the Systems office's purge of logically deleted auth records from the ORBIS database (166,577 records). Any auth record deleted from this time forward will not truly be deleted from ORBIS and will migrate to Voyager as an undeleted record. We don't want this to happen!

Upcoming Authority Work Restrictions in ORBIS. From May 13th-June 5th, staff will not be allowed to create new auth records in ORBIS (by keying, CLARRing, or DERIving). They will also not be allowed to update or overlay existing auth records (by keying, AOVLing, and CLARRing). Any changes to the authority file made after May 13th will not migrate to Voyager. NACO auth record creation and updating can continue as usual in RLIN. Staff should make printouts of local-level updates (090, 644, and 646 decisions) needed on existing auth records and key these updates once we are working in Voyager.

Deduping of Auth Records in LTLC. The Systems Office will be deduping the authority file in LTLC. If two (or more) records share the same LCCN, the record with the highest LTLC # (which should be the most recent version) will be retained. This deduping will not take care of all of our problems--for example, auth records deleted from the national file will not get deleted in LTLC (and will inevitably still be used by staff in new cataloging--but it is a big improvement and helps to make our Resource File much more authoritative.  [Note: the deduping was subsequently performed on May 1st and resulted in the deletion of 53,419 duplicate auth records].

Migration of N/Y range auth records without 010s. The Systems office asked if we wanted to migrate N/Y range auth records that lack 010s to Voyager (18,827 records). These records are primarily locally-created auth records from before we became a NACO library. Some of them are duplicates of national-level records that have since been created in a different form; we delete these records as we encounter them. Others are unique and have no counterpart in the national file. Although there would be some advantage to purging these records and having an authority file in Voyager that contains only national-level records, it was decided that the records contain too many valuable cross references and too many series analysis and classification decisions for us to purge them. The records in Voyager will carry the old N/Y range ORBIS key in the 035 field. However, the Voyager auth loader does not match on 1xx, only on 010, so these records will unfortunately not get overlaid if another NACO library creates an auth record with the same heading.

"ID Catcher" Program. The Systems Office is writing a new program called "ID Catcher" that will serve as a replacement for the current practice of adding 927 and 928 fields to export bib records to MARS and to the utilities. The "ID Catcher" window sits on top of the Voyager window. You can move it to wherever you want on the screen. It contains the Voyager ID for the record you currently have displayed and buttons for sending to OCLC, RLIN, and MARS. You just make sure the appropriate buttons are selected and click "Send" once you have finished cataloging. As soon as you do this, another window pops up called the "Send History" window. This window contains a list of all the records that you have flagged for export during a particular session and where they are going to be sent. The feature that is still being developed is the part that warns you if the record has previously been sent to MARS but not yet received back and loaded. We will be moving to full record overlay for our MARS processing in Voyager so it is important that we not update a record before we have received it back and loaded it. Otherwise our changes will be overlaid. We do not know exactly what this warning will look like, but we hope it will be distinctive! Another "plus" of the program is that it allows you to copy the bib record ID # and paste it into another application (e.g. an e-mail message or a Word document). You cannot currently do this from the Voyager window.

Voyager Authorities Workshops and Documentation. The Committee will need to revise existing authorities documentation and develop authorities workshops as needed. Possible workshop topics include heading validation, identifying cancelled auth records in LCDB, and importing auth records. The Committee had also previously identified the need for a basic series workshop.

Site URL:
Comments to: Manon Théroux  <>

© 2002 Yale University Library

Top of Page