Expert Community Experiment: Guidelines for Experts

Expert Community Experiment: Guidelines for Experts

February 2009. Jay Weitz, Consulting Database Specialist

Background

OCLC's Expert Community Experiment is intended to provide Connexion users who have a Full Level authorization or higher more flexibility in making changes to WorldCat master bibliographic records. Maintenance of WorldCat will be shared more equally between OCLC staff and member libraries. The additional capabilities provided by the Experiment are a powerful expansion of those that have been available especially through Database Enrichment since 1991.

OCLC hopes that these changes will result in more corrections and additions to master bibliographic records and more timely actions to correct record problems. The Experiment is inspired by longstanding requests from members of the OCLC cooperative to be able to do more -- and more immediate -- upgrading of bibliographic records in WorldCat. The Experiment allows us to test a "social cataloging" model involving the existing community of cataloging experts who have built WorldCat record-by-record over the past nearly four decades. Interest in such an experiment has grown especially with the launch of WorldCat Local and with the popularity of such other socially cooperative ventures as Wikipedia.

Authorizations

Any Full-Level authorization or higher (including CONSER Regular, CONSER National, Regular Enhance, National Level Enhance, NACO Regular, NACO National, and Agent) will be able to take part in the Experiment. There is no need to apply for any special authorizations. All existing Enhance, CONSER, and NACO authorizations will continue to work as they always have but will also have all of the additional capabilities allowed within the Experiment.

Basic Principles of the Experiment

The overriding principle of the Experiment is: "First, do no harm." Please use the same care in editing an existing master record as you would use in creating a new record.

• NEVER remove correct and accurate information from a master record (such as classification numbers or subject headings) simply because your institution does not find it useful.
• NEVER change the basic nature of a master bibliographic record into something different.
• AVOID including local data or local practices in master bibliographic records.
• Be cautious about changing bibliographic records with a different Language of Cataloging (identified in field 040 subfield ‡b). Bibliographic Formats and Standards, Section 3.10, "Parallel Records for Language of Cataloging" http://www.oclc.org/bibformats/en/specialcataloging/default.shtm#BCGBAEHC) has details on working with these records.

A second overriding principle of the Experiment is: "If in doubt, DON'T."

• Do not replace a record solely to change an element that is a matter of cataloger's judgment.
• Consider data within the context of the whole record, rather than in isolation. If you can reasonably resolve contradictory information within a record, please do so, but try not to jump to conclusions.
• Please behave responsibly concerning each other's records. Remember that each library that replaces a record input by another library appears, by virtue of the 040 field, to have some responsibility for the content of that record.

Bibliographic Records Included and Excluded in the Experiment

The Expert Community Experiment allows members of the OCLC cooperative with Full-Level cataloging authorizations and higher to make additions and changes to almost all fields in almost all records. There are a few relatively limited exceptions and a few additional details:

• PCC records -- both BIBCO and CONSER records -- are EXCLUDED. These records are identified by one or more of the following field 042 authentication codes: isds/c, lc, lcd, msc, nlc, nsdp, nst, pcc.
• LC records that are NOT coded as PCC records are INCLUDED.
• CIP records (Encoding Level "8") that are NOT coded as PCC can be changed, but the Encoding Level must remain at "8"
• Institutional Records (IRs) are EXCLUDED.
• Records in the Hand Press Books database are EXCLUDED.
• Records with Encoding Level "E" are EXCLUDED.

One important goal of the Experiment will be determining whether these restrictions make sense, whether they are too broad or not broad enough.

Bibliographic Data that Cannot Be Changed

Most fields within an eligible master bibliographic record can be changed, with the following EXCEPTIONS:

• System-supplied data (040 subfield ‡c, Entered).
• Fields 019, 029, 066, 850, and 938.
• Field 042 (except that, when upgrading an Encoding Level "3" record, you can remove field 042 with code "dc").

Note also that master bibliographic records CANNOT be deleted from WorldCat as part of the Experiment.

Mechanics

For users of the Connexion client, you can replace a record without locking it first and you can do replace transactions in batches if you prefer.

For users of the Connexion browser, you must lock a record before you begin to edit it, and then replace it.

Some Important Points About the Lock and Replace Process:

• At least one field retained in the master record MUST be edited before you can do a replace.
• Your OCLC symbol automatically appears in 040 subfield ‡d after replace.
• "Replaced" date changes.
• A Replace transaction DOES NOT set your holdings.
• OCLC requests that work on locked records be completed as expeditiously as possible. This affords other users access to your improved record right away and allows OCLC automated database quality software to work more effectively. Locked records cannot be processed by Duplicate Detection and Resolution software, automated authority control software, or database scans
• Please see Connexion Help for your WorldCat interface for details of record expiration in the save file.

You may want to re-familiarize yourself with the sorts of local information that can be present in a record at the point of replace but that is not added to the master record. See the section "Replacing Records with Local Information" in Bibliographic Formats and Standards: http://www.oclc.org/bibformats/en/quality/default.shtm#CIAIGHAI

Credits

At least for the duration of the Experiment (expected to be approximately six months between February and August 2009), the existing structure of bibliographic record replace credits (including CONSER, National Level Enhance, Regular Enhance, Database Enrichment, and Minimal Level Upgrade) will remain unchanged. If you received a credit for a replace transaction prior to the Experiment, you should continue to receive the same credit.

If you were not previously authorized to replace a record or would not previously have received a credit for a replace, you will not receive a credit as part of the Experiment. OCLC will be monitoring replace activity performed as part of the Experiment under a new product code (ONT6390). Later in the course of the Experiment, OCLC will determine lump sum credit adjustments where appropriate. The overall idea has been to keep participating institutions "whole" with respect to the credits they have previously been eligible for.

OCLC will be re-evaluating the whole replace credit structure both during and following the Experiment and will keep users informed about changes.

Sharing Information on the Experiment

OCLC hopes that members of the cooperative participating in the experiment will share their experiences on the OCLC-CAT discussion list (OCLC-CAT@OCLC.ORG). In addition, questions on the Experiment may be addressed to ASKQC@OCLC.ORG.

The Future of the Enhance Program

Clearly, the Experiment changes the face of the Enhance program. OCLC has been talking with Enhance participants about this since the ALA Annual meeting in Anaheim, California, in June 2008. In these discussions we have come up with some possible expansions of Enhance capabilities in the future, such as the ability to merge certain categories of master records, maintenance of headings, and the like. OCLC will continue to have these discussions with our valued Enhance participants during the course of the Experiment and afterwards.