Two topics were on the agenda, the proposal to add records created by OCLC as part of the e-Serials Holdings Pilot to the CONSER file and an update on the development of the access level record for serials.
CONSER Program Report
Saint Louis University Pius XII Memorial Library and Health Center Library was welcomed as a new associate level library. Shana McDanold is the new associate level representative. Serials Solutions also has applied and will be welcomed as a new affiliate member doing maintenance on CONSER records.
OCLC record discussion
Les gave an introduction. CONSER has been considering how it can increase coverage of e-serials in electronic packages for sometime. From the first CONSER group that looked at controlling titles in packages in 1998 to the three PCC aggregator task groups that followed from 1999-2004, there has been recognition that the separate record approach is one of several methods libraries use to control titles in packages. There also was recognition early on that libraries do not have the resources to systematically create separate e-serial records for titles in packages, especially by hand. At the same time, for those libraries using separate records or vendor products that provide metadata about titles in packages, incomplete coverage of package titles and the lack of fuller metadata have been a problem.
The aggregator task groups worked with vendors to develop record requirements for the automated creation of records based on existing print records. In conjunction with the discussion of the aggregator neutral record e-serial record in 2003 [ discussion from OpCo meeting summary, 2003], CONSER members discussed categories of machine-derived records appropriate for inclusion in the CONSER file (e.g. full text titles versus article databases) and ways to identify them as machine derived. The work of the third task group on aggregators [ final report, August 2004], included the development of a macro by Robert Bremer (OCLC) to clone existing print records and provide a mechanism to authenticate resulting e-serial records. Several libraries experimented with the macro and several hundred e-serial records created from cloning existing print records were added to the CONSER file with the macro. (The macro and the data element set developed for extraction or modification in the cloned record were later used to derive records in the eSerials Holdings Pilot.)
In 2004 the CONSER Summit included the recommendation to increase coverage within the CONSER file of titles in e-serial packages. The Summit included many representatives from vendors and other subscribers to the CONSER file.
During the summer of 2005, representatives from OCLC and LC’s Serial Record Division discussed OCLC’s e-Serials Holdings Pilot and the possibilities of adding these records to the CONSER file. CONSER members discussed the proposal to authenticate the records in October 2005. Les compiled questions from the CONSER conversation about the project and asked OCLC representatives to consider them before the at-large meeting. Bill Carney & Glenn Patton provided the following responses.
1) Please explain the nature of the holdings added for e-serials in WorldCat. From LC's comments on the project: "Do e-serial holdings reflect "master" holdings, in the sense of an idealized run of all published holdings (as opposed to the holdings or a particular institution), or local institutional holdings?"
A: Nature of the holdings: The eSerials Holdings Pilot process is setting holdings at the journal title level, i.e. MFHD level one. The holdings are set for each pilot library against 31,000 MARC bibliographic records. A future version of the service will include setting detailed holdings, i.e. MFHD level four.
2) How do the e-serial bib records fit into the bigger picture of Open WorldCat or other OCLC services? Besides being able to add holdings and distribute bib records through CDS, do the bib records serve other purposes for current or future OCLC services?
A: In the larger picture, OCLC services are not segregated, e-serial holdings are registered against WorldCat bibliographic records, which are discoverable across OCLC’s service platform. Libraries are able to link to the items through services such as FirstSearch and Open WorldCat if they have registered their OpenURL resolver in the FirstSearch administrative module or in the OCLC OpenURL Resolver Registry. Pilot libraries have indicated increased operational efficiencies in ILL as their staff discover holdings for e-serials in WorldCat and can fill internal requests from their own collection.
3) Some mentioned that the timeframe for adding the records to the CONSER file seemed precipitous. Glenn Patton mentioned in November that waiting too long to add these records to the CONSER file would make them a less valuable resource. Could you address this for members?
A: For those records created from print records with the first large group or blob of records in June, and for records created later on an ongoing basis, there is a danger that changes made by CONSER members or OCLC TechPro staff in the print record will not be reflected in the e-serial records. If there is a large time lag in subsequently authenticating these records, there is increased risk of the print and online records being out of synch in the CONSER file.
4) Could OCLC address the duplicate resolution situation? LC's reply summarizes some of the concerns, including that staff are still running into a number of duplicates, that there is a need to extend the timeline to assure duplicates are caught before authentication and the need to assure that there is a mechanism for catching future duplicates created through automated processes.
A: There is an ongoing effort to identify and resolve duplicates. Duplicates created on the basis of existing e-serial record were caused by the lack of code “s” in existing source records, which would have identified it as an existing e-serial record. Records without print ISSN sometimes caused a problem. Additional checks and workflow have been put into place to help avoid future duplicates like this. Acquisition of the assets of Openly Informatics and use of their1.2 million record knowledge base is expected to improve identification of target records.
5) What has been the feedback from CONSER file subscribers about the prospect of adding the e-serial records to the CONSER file?
A: Four vendors are partnering with OCLC in the e-serials holdings project, Serials Solutions, TD-Net, ExLibris, and Ebsco, they are subscribers to the file and have been very positive about potential addition of the records to the CONSER file.
6) OCLC subscription service vs. CDS distribution. Some have asked about the
possibility of companies receiving these records through an OCLC bib record
subscription service rather than through distribution of the CONSER file.
A: OCLC doesn't have a ready mechanism to deliver these records directly to vendors. We could probably develop such a mechanism, but shouldn't we use the existing record flow to assure that all subscribers to the file, including RLG receive the records? Also as a not for profit, OCLC needs to be very careful in its arrangements with for profit companies.
7) There were questions about how maintenance of bib records and holdings information will be handled, including keeping up with title changes especially after the e-serial pilot ends. For bib records, some made the point that the e-serial records won't receive the same type of scrutiny that member authenticated records receive, in terms of libraries catching title changes for serials they hold and have cataloged. What would be the maintenance model for these titles?
A: Traditional cooperative maintenance in CONSER would occur. OCLC TechPro staff are also relied upon to close out records and creating new records.
8) Is there any way to get a general idea of the types of resources represented by the e-serial records? From LC's comments: "Staff associated with projects involving digitization of serials originally in print, e.g., the National Digital Newspaper Program, also wondered what types of material were included in the records to be added to the CONSER file. Understanding the criteria for inclusion, and thus the likely overlap, would help those involved in such projects at LC and other institutions. In particular, it would allow them to determine whether they need to adjust plans, processes, or work flows."
A: The composition of the record set is completely dependent on the subscriptions that libraries are tracking through their service providers, which is directly reflected in the holdings data sent to OCLC from partners and libraries. As libraries add e-serials to their collection, service providers add the journal to their knowledge base, and holdings are sent to OCLC. OCLC’s recent acquisition of the assets of Openly Informatics will enable OCLC to keep WorldCat up to date with the latest list of available e-serials.
Questions and comments from the audience:
There was a question about how the ISSN of the print record ends up in the 022 $y for these OCLCS records [OCLCS is the 040 code for the project].
A: CONSER practice is to record an eISSN if available and known in 022 $a on the e-serial record and the ISSN of the print if known is recorded in $y of the e-serial record to provide both for searching. The project followed this practice in creating the records.
Straw poll: How many use the single record approach? Holdings added as a part of this project won’t show for the single record approach, most in the room indicate they are using the dual record approach to some degree or another. Some attendees are getting e-serial records from Serials Solutions
Regina Reynolds raised the concern about e-serial records created by the project for resources that are integrating resources in online format, though they are serial in the print version. Also, she and her staff have noticed that e-serial records are sometimes created for titles where only a few scattered articles appear in electronic format, and a real online version does not exist.
There was concern expressed if identifying these records with a new 042, “oclcderive” that it may more difficult to interpret CONSER records.
Comment: We seem to be discussing two different things, WorldCat records and vendor records. Maybe best to leave the records in WorldCat and provide a separate vendor product.
Comment: CONSER needs to maintain its relevance in a changing world. If it can provide useful metadata for gaps in vendor products, it should take advantage of that opportunity. In this case, CONSER can help supply metadata that is fuller and in many cases more authoritative than the minimal title/ISSN that some companies have for gaps in aggregator collections.
Les commented that he agreed with this last statement and said he sees this proposal as an opportunity to leverage existing CONSER records to provide a service that would be useful to subscribers of the CONSER file. Based on discussion of the issue in October and during the at-large discussion, most members were comfortable adding records created based on existing CONSER print records to the file. Les suggested a middle ground approach: perhaps we can begin with loading the records created based on existing CONSER print records (about 11,500) and monitor how the process goes.
Access level record for serials
Regina Reynolds and Diane Boehr (NLM), co-chairs of the LC/PCC group working on this project gave attendees an update. The idea began with discussion of a presentation that Dave Reser (LC) gave at the 2005 CONSER/BIBCO Operations Committee meeting on LC’s access level record for non-serial electronic resources. Like LC’s project, the goals of the LC/PCC serials project are to increase cataloging efficiencies and better meet user needs by providing the necessary elements that support resource discovery.
The access level record for serials supports FRBR user tasks, conforms to AACR2, LCSH, MARC 21, and is compatible with records distributed to MARC record subscribers through CDS.
The cataloging guidelines developed by the project are very well synchronized with many of its principles in RDA, including a list of mandatory elements, a focus on transcription for description, and reliance on controlled access points. The access level record for serials probably can be implemented before RDA is complete without conflict. The access level group has been providing input directly into the development of RDA through the JSC.
Highlights from the cataloging guidelines include the recommendation that all 362 fields will be unformatted to better convey to users the issues with which a serial began or ceased rather than imply that the library actually holds the first or last issue. The group found that serial “distinguishing” uniform titles interfere with FRBR displays, so will recommend that distinguishing uniform titles be minimized. In the future, recommendations will be made to change the current practices for creating uniform titles for translations and language editions.
Only some notes are required by the guidelines, source of title, description based on, latest issue consulted, if applicable. Other notes are of limited use or are not required. The group will consider proposing a MARC 21 change to suggest better linking field indicators so that 580 fields don't need to be used for as many complex relationships. The principle behind this is to eliminate redundancies in the record.
The “title change analysis chart” has been created to provide an easy, reliable reference for making major/minor change decisions. The guidelines reflect an increased reliance on authority records and guidance for establishing corporate headings, preferred solutions, and “if in doubt” help with forms of headings subordination, etc.
Pilot testing of the record elements and the cataloging guidelines is planned for March-April 2006 will involve 10 or so libraries each with teams made up of cataloging, reference, and acquisitions staff.
Will training be provided for catalogers before testing? Yes, this is something the group has discussed and is planning for.
Concerns were raised about changes in practice with added entries and uniform titles.
A: There is nothing changed about adding access points. With uniform titles, exceptions need to be made for mono series and generic titles, some practices may not be consistent with AACR2.
For the future to think about: will this be a PCC standard or more stand-alone and PCC endorsed?
There is a need to consider record exchange and compatibility with international records. The first decision in the title change analysis chart focuses on a change in meaning/subject matter that requires new subject headings, what about non-LCSH? Maybe there is a need to look at subject headings in a larger, international context.