Access to Serials within Aggregator Databases John Riemer 10 Dec 1998 18:36 UTC

This message is being cross-posted to AUTOCAT, CONSRLST, LIBREF-L, PCCLIST
& SERIALST.  Apologies for the duplication.

SURVEY ON PROVIDING ACCESS TO SERIAL TITLES WITHIN AGGREGATOR DATABASES

I.  Purpose.  To determine the need for access to journal titles in full
text aggregator databases, the most desired form of access, and the level
of willingness to cooperate on providing such access.

II. Background.  A growing number of libraries are subscribing to various
aggregator databases, such as ABI/Inform, Lexis/Nexis, and UMI Proquest,
which contain the full text of thousands of electronic journals.  In most
(all?) cases, these journals have print equivalents.  The databases vary
in coverage and in the format in which journals are presented within them.
In many cases, access is only provided at the article level, making it
impossible to describe the journal title according to traditional AACR2
methods.

In one sense, these databases are similar to microform sets, for which
librarians have collaborated in creating sets of records.  However, in
another sense they are very different.  While microform sets are stable,
aggregator databases are very dynamic.  Titles are added and dropped with
little notification to the subscriber.  It is even more challenging to pin
down & stay up with range of volumes available for a given title.

The CONSER Program first encountered the problem in the late 1980's with
what we called "mega" CD-ROMs discs that could contain up to 2000 titles.
We discussed ways in which to provide access but the fact that the title
coverage was so unstable kept us from taking any action.  Now we are faced
with the online equivalent and we can no longer ignore the issue.  The
subject was discussed at ALA in June 1998 and a small working group was
formed to explore the options and make recommendations to the Program for
Cooperative Cataloging's Policy Committee.  The group is chaired by Ruth
Haas (Harvard);  other members are John Riemer (Georgia), Jeanne Baker
(Maryland), Karen Calhoun (Cornell) and Jean Hirons (LC, ex officio).
This group is seeking advice from the library community on desired forms
of access and ways in which we can contribute to the effort to provide
such access.

Results of this survey and possible action for CONSER will be addressed by
John Riemer, Karen Calhoun, and others at ALA's Cataloging Management
Discussion Group meeting in Philadelphia.

III.  Methods of access.

III.1.  Lists of titles on institutional web sites.

This is the method that many libraries have been using.  While some seem
to be satisfied with this approach, others are finding the growing number
of titles unwieldy and are looking for more traditional access through the
library's OPAC.

III.2.  CONSER single record approach.

This involves using the print (or CD-ROM) record to note the availability
of the online version.  This method has been very widely used for a number
of electronic journals, particularly those that are print equivalents.
The drawbacks to this approach for aggregator databases are that the print
may or may not be held by the library and it would not be easy to use such
records for distributed record sets.  The primary advantage is that it
reduces the number of records in the OPAC for any given journal title.

III.3.  Separate records for titles in one or more databases

Having a separate record for the journal as contained in the database (or
multiple databases perhaps) is appealing in that the records could be
bought as sets.  Whether or not such records could include more than one
database (using field 773 Host Item entry) is a matter for discussion.
For example, having a single "aggregator database" record on OCLC that
contained all databases would be desirable if OCLC could customize the
records for sale to individual libraries.

There are issues related to how to create such records since one cannot
really view the journal as an entity.  However, since most titles exist in
print, a brief record could be created using the print record data and
containing the authoritative title, ISSN, coverage, and database in which
it is contained.  If this approach is deemed most useful, CONSER could
work on a standard set of data elements to be used in these records.

OCLC currently has a couple of projects, WorldSets and CORC that result in
the creation of separate records.  The latter can be based on the Dublin
core but does not involve the creation of metadata within the resource
itself.

Publisher-supplied metadata within the resource that could be mined to
create a record is another source of separate records and worth exploring.
Libraries could collaborate with publishers in this effort.

III.4.  510-field approach.

This involves viewing the aggregator as an indexing and abstracting tool.
It would be more convenient to add fields to an existing record than to
create additional records.  Coverage dates could be closed out as
warranted.  As with the single-record technique, use of the 510 field(s)
would presume ownership of the print or other tangible equivalent;
additional limitations are the current 50-field limit and overall record
length and complexity once that barrier is removed.  (The challenge of
maintaining the existing 510 fields has not yet been met either.)

III.5.  Holdings records attached to print record.

For some institutions, the most desired approach may be to create a
separate holdings record attached to the print record indicating the
availability and coverage of the title in the database.  This presumes
that the print is held and will be retained.  It also seems to be a purely
local option at the moment.

IV.  SURVEY  (Deadline January 8, 1998)

***To respond to the survey interactively (the preferred method), please
go to http://lcweb.loc.gov/catdir/pcc/consersurvey.html
to complete and submit the form available there.  Alternatively, you may
fax your responses to the questions below to John Riemer at (706)542-0591.
Please consult with your public/technical services colleagues in
formulating a response, and send just one survey from your institution.
Thank you for your help.

IV. 1.  Does your institution currently license full-text e-serials
via aggregators such as Lexis-Nexis, ABI/Inform, etc.?
      A. Yes
      B. No

IV.2.  Which access method are you using now? (Check all appropriate)
     A. Lists of titles on web sites.
     B. Single record approach
     C. Separate records
     D. 510 field added to print record
     E. Holdings attached to print record
     F. None of the above
     G. Comments

IV.3. Which access method would you most like to employ in your
institution?  (Check one item)
     A. Lists of titles on web sites.
     B. Single record approach
     C. Separate records
     D. 510 field added to print record
     E. Holdings attached to print record
     F. None of the above
     G. Comments

IV.4.  My institution would be willing to: (Check all appropriate)
     A. Work with CONSER libraries to create and maintain sets of records
          using either the single or separate record approaches .
     B. Explore the possibility of working with publishers to create metadata
     C. Purchase sets of records
     D. Comments:

IV.5.  Further comments or suggestions

John J. Riemer
Assistant Head of Cataloging
University of Georgia Libraries
Athens, GA  30602
(706)542-0591 voice
(706)542-4144 fax
jriemer@arches.uga.edu