Rev. Dec. 2, 2002; approved by the PCC Policy Committee Nov. 8, 2002
Written by Jean Hirons
While a majority of respondents to the CONSER survey selected Option B (about 2 to 1), it is obvious from the responses of those favoring Option A that there is a real need to maintain some type of aggregator information in records to make them useful. The following is a compromise proposal based on the comments received from the survey, discussions with various CONSER members and Robert Bremer. It aims to achieve the following goals:
- Provide a clean record that can be used by all
- Provide an efficient means for collapsing existing records to create the single CONSER record
- Minimize the amount of CONSER cataloger maintenance needed for these records and maximize automatic maintenance by OCLC
- Minimize the amount of local editing needed to use the records
- Make the records identifiable for libraries or companies wishing to produce record sets
- Make the records findable under variant title access
- Maximize use of the ISSN
- Do not strip useful data from existing records.
Important caveat: There are two primary types of aggregations or packages. Those that are journal-based (i.e., a distributor of the complete journal in its electronic form; and those that are article-based (i.e., large databases, such as ProQuest, that contain all or selected articles from a journal). Under the following proposal, CONSER will add/maintain data for journal-based aggregators only. However, serials management companies or libraries could use the records to locally add data for the article-based aggregations.
Features of Option B Plus
- A single record for all online versions
- Based on the journal as found on the publisher's Web site, when possible
- Containing the URL of the serial as found on the publisher's Web site when this forms the basis of the description
- ISSN of the print in the 022 $y (as a short term solution)
- 246 fields for variant titles found on aggregations with $i "Distributed as:"
- The 856 of known journal-based aggregations (i.e., those from existing records to begin with)
- CONSER should work with the serials management companies to identify and authenticate titles that are not in the CONSER database
- CONSER and NSDP should work with serials management companies and ISSN International Centre to supply missing ISSN and provide a more long term solution to accommodate use of ISSN for multiple formats in these records
- CONSER will work with serials management companies to educate aggregators on the presentations of serials in their collections
- PCC task force to work with ILS vendors to develop means for bulk load and batch processing of these records
This proposal is a compromise that aims to maximize the usefulness of CONSER records, while minimizing maintenance. After reading many comments on over 100 surveys received, it is clear that one record is desired and that CONSER should find a way to minimize both the local editing that would be needed to use that record as well as the need for CONSER maintenance. But it is also clear that for those not using a serial management company, completely stripping the OCLC records of any aggregator data would be a disservice.
An initial draft of Option B+ was sent to the CONSER E-Serials specialists, which called for use of the 773 and no 856 fields other than the publisher's. Discussion on what a publisher is, which 856's to retain or delete, and the difficulty of doing this on a selective basis when collapsing records, have been influential in changes to this proposal.
Discussion on elements of the proposal:
246: The 246 fields could reside in the record regardless of whether they appear on the package that a library has (i.e., there is no harm).
ISSN: The print ISSN is needed by many libraries and it is heartening to see that steps are being taken to remedy this situation. The proposal calls for short and long-term solutions.
856: The 856 field would contain only the URL. No coverage information or special notes are needed. The URL contains the name of the aggregator and is sufficient for access. The URL should be a non-local URL that could be used by any user whose library subscribes to the aggregation.
There has been much discussion of the 856 since the initial draft of this proposal was floated. There is much confusion over which would be eligible, which wouldn't, and how catalogers would make this determination. A suggestion to remove all 856 fields would not serve the needs of the ISSN center. Robert Bremer has also noted that we can remove the data later, but it would be more difficult to add it, and in the future there may be resolution servers that might reduce the number of URLs needed to one. CONSER catalogers, in some cases, are taking great pains to create useful and universal URLs and it seems unwise to remove this data from the records. Thus, the proposal to retain the URLs as they are currently in the records and to add new ones as new packages are received.
Maintenance: Many of the comments from non-CONSER respondents in favor of Option B were concerns for the burden of maintenance on CONSER. Discussions with Robert Bremer have made it clear that adding, deleting, or changing URLs for an entire package is quite doable by OCLC and need not be a burden to individual CONSER catalogers. If there are titles added/dropped within packages, perhaps the serials management companies could notify OCLC or certain CONSER libraries for maintenance. There seem to be many possibilities here. At present CONSER is creating, maintaining, and collapsing separate records! Maintenance of a single record will be much simpler.
Database clean up: Under this proposal a single record would be selected, and other records would be collapsed. The 856 fields from the other records would be retained but 5XX and 7XX fields that pertain to individual aggregators would be stripped. The clean up will be accomplished by OCLC with advise from CONSER catalogers.
Following are possible scenarios for how these records would be used by those who preferred Option A or Option B, and by CONSER libraries.
Libraries that selected Option A:
- Would identify the records via the URL, title or ISSNs
- Would use OCLC macros to add local data
- Could decide to delete, suppress, or retain unwanted 856 fields
Libraries that selected Option B: (most are using a serials management company)
- Serial management company would use the record as the basis of their record, deleting and customizing information for local use
- Or: A library would identify the record set needed, add/delete data as above
The primary disadvantage for those who selected Option B would be the potential need to both add and delete data. However, if it is determined that the existing data in the records is not misleading to users, it could be retained, with local data added.
CONSER libraries: (the following will be done in phases)
- Would add new records based on serial as found on publisher's Web site
- Would add new 856/246 data to existing records
- Would maintain bibliographic description
- Would work with OCLC on the clean up of the existing database (see below)
- Would work with serials management companies to identify titles needing authentication
- As appropriate and possible, would help supply ISSN
Issues needing to be resolved
There are many issues that need to be addressed, some more critical than others. Below is a list of the issues, in no particular order along with a suggestion for who will address them and when.
- Uniform titles. Two issues relating to the need for a more collocating uniform title.
- Should the qualifiers of the print and online match with the addition of "online"? (e.g., if print has a place qualifier, use that place in qualifier of online, regardless of whether the place is the same or ever associated with the online).
- Should the 130 title in the online record match the 245 title proper of the print? This may be important when using an aggregator as basis of description and titles vary.
Suggested task members: Judy Kuhagen (LC), Regina Reynolds (LC), David Van Hoy (MIT), George Wrenn (UCLA), Kevin Randall (Northwestern), Frieda Rosenberg (UNC)
- Source of the description. There is general agreement to use the publisher's Web site as the preferred source of description. We need to agree on what will be the second and third choices.
- Use the print as the second choice. Most survey respondents indicated this. Aggregator would be third choice.
- Or, use the aggregator being cataloged as second choice; a few CONSER catalogers have indicated a preference for this.
Suggested task members: Edith Gewertz (NYPL), Jennifer Edwards (MIT), Valerie Bross (UCLA)
- ISSN issues. Many respondents indicated problems with the ISSN. Many noted the need for the print ISSN in the records because some of the aggregators use the print. Regina has agreed to the use of the print ISSN in $y of the 022 as a short-term solution with the understanding that a better long-term solution will be investigated.
- Explore a long-term strategy to supply ISSN for multiple formats in CONSER records, including the development of a MARC proposal, if applicable.
- Work with serials management companies and the ISSN International Centre to identify and add missing ISSN to CONSER records for online serials
Suggested task members: Regina Reynolds (LC), Jean Hirons (LC), Becky Culbertson (UCSD), Diane Boehr (NLM), Kevin Randall (Northwestern), Kathleen Dougherty (NAL).
- Database issues. There is a clear mandate to clean up the existing records, merging the multiple records into one. A small group should be formed to propose an implementation strategy. Issues to be addressed include:
- Choice of records (use of the NSDP record, when possible, will probably result in the best first choice);
- What to retain, delete, change;
- How much can be done automatically and how to handle manual changes;
- Deauthentication of the CONSER records to be deleted
Suggested task members: Robert Bremer, John Levy, Les Hawkins, Becky Culbertson (UCSD), Edith Gewertz (NYPL), Steve Shadle (Washington), George Wrenn (UCLA law), Fieda Rosenberg (UNC).
- Non-CONSER OCLC contributions. While the CONSER database is discrete, it resides within OCLC. For the benefit of libraries using OCLC, it would be desirable that OCLC members not continue to add new records for aggregations but find and add their holdings to the existing CONSER record. However, there may be a desire on the part of some OCLC members to be able to add data to these records for the benefit of all (i.e., additional 246 or 773 fields).
- Work with OCLC on a policy for non-CONSER contributions
- Open CONSER Enhance membership to libraries who would like to help maintain online record
Suggested task members: Glenn Patton, Jean Hirons
- Improving CONSER coverage. A number of respondents commented on the fact that better coverage within CONSER would greatly enhance the process. With this in mind, CONSER should work with the serials management companies to develop a strategy for identifying and authenticating records for titles not in the CONSER database.
Who: Bob Wolven, Jean Hirons, David Fritsch, Peter McCracken
- Educating aggregators. Aggregators have a tendency to ignore title changes or use a latest entry approach when presenting their titles. It has been suggested that CONSER spearhead an effort to educate the aggregators on title presentations. CONSER might wish to enlist the services of the serials management companies in this task. Information forums at ALA would be appropriate.
Suggested task members: Regina Reynolds (LC), Charlene Chou (Columbia), James Castrataro (Indiana), Ann Ercelawn (Vanderbilt), David Fritsch (TD Net), Peter McCracken (Serials Solutions)
- Automation aspects of using the CONSER records. Helping libraries develop the best means of using the records will be an added benefit that PCC can provide. Some libraries have been able to develop in-house mechanisms, others are having more problems, some due to the limitations of their ILS. Working with ILS vendors to develop mechanisms for bulk loading and batch processing the records locally would be useful.
Who: Adolfo Tarango and PCC Task Force on Journals in Aggregations
Suggested Implementation Time Frame
|November:||Approval by PCC; discuss database cleanup|
|December:||Make preparations for discussion at ALA
Begin work on Priority 1 items
|January||Hold discussion at CONSER At-Large and other meetings.|
|February:||Groups should begin work on Priority 2 issues for discussion at CONSER meeting
Begin database cleanup
|May:||CONSER discussion and approval of cataloging-related issues
Implementation for new cataloging begins
|June-July:||CONSER documentation prepared|
|September:||Begin work on priority 3 issues|
- Original Proposal: Single Records for Online Versions of Print Serials May 2002
- CONSER discussion of this proposal at the May 2002 Operations Meeting
- Discussion of the proposal during CONSER At-Large Atlanta (June 2002)
- Survey: CONSER Proposal to Create a Single Record for Multiple Distributors of an Electronic Journal (August 26, 2002)
- Survey Results: Single Record for Multiple Distributors of an Electronic Journal (Sept. 23, 2002)