The morning meeting initiated our discussion of best practices for electronic resource cataloguing. This is an extension of the best practice concept that was in place, on other topics, back when Prospector began. We dealt with usage and function of the ISSN (022 field) and URLs (principally 856 field).
ISSN discussion
An electronic version of a journal *should* carry a unique ISSN and not the one
for the paper copy. In the real world, variety is rampant. Some journals have
received e-ISSNs and some have not. Some bib records have these ISSNs and some
do not. Vendors use different link-to-holdings algorithms. Some use an ISSN
index, thereby picking up the e-ISSN only if the library has entered it in the
022|a. Some look at the ISSN in the record, sometimes reading all instances,
sometimes reading only certain subfields, sometimes reading only the first ISSN
listed. Apparently vendors are sometimes unable to tell us how their algorithms
work.
The two basic ISSN concerns are 1) the effect of ISSN usage on the overlay
algorithm in Prospector, and, 2) the effect on link-to-holdings by various
vendors of ISSN usage and placement in the bib record. In regard to overlay:
Prospector overlays on the 001 and the 245|a. If no match is found, Prospector
seeks an 022/title match. A phone call to III indicated that the ISSN overlay
works off the ISSN index, which contains only the 022|a (we do not index the
ISSNs in the |y or |z). III will call George back to verify for sure that it
runs on the index and not the bib record. Thus an ISSN for the incorrect format
(i.e., an e-ISSN on a paper record, or a p-ISSN on an electronic record) in the
022|a can result in an incorrect overlay in Prospector. Indexing the 022|y or
022|z would result in worse overlay problems.
For purposes of overlay, the issue is solely whether the ISSN is
indexed, not whether it is in the 1st or 2nd instance of the 022 or whether an
022 has a single or multiple instances of the |a.
Potential best practice recommendation for correct overlays: be sure the
correct ISSN is in the 022|a; on a single record for multiple formats be sure
the 022|a is for the primary format being catalogued; do not index the |y and
|z. Obviously we can follow this practice only if the vendor supplies separate
ISSNs.
Regarding linking-to-holdings by ISSN matching, we can do nothing about the
various vendor algorithms in use; we cannot always get a separate e-ISSN. All
we can do is, using the best data available, enter the correct ISSN in any bib,
or both ISSNs if we have a single record for both the paper and electronic
versions.
Potential best practice to facilitate linking to holdings: Make the 022 |a, |y,
|z available for linking by vendors' open-URL services. Be sure the first
instance of the |a is the ISSN for the item being mainly catalogued.
URL discussion
Regarding use of the 856|z: there is great variety among Prospector libraries
in the use of the 856|z to convey information. It seems clear that, at this
point, |z usage will need to be a local choice, with only limited Prospector
advice.
Possible best practice on 856|z usage: use the 856|z to: 1) identify sites or
resources related to the title being catalogued (indicators 42) 2) identify
partial sites or resources, such as abstracts or limited years (indicators 41)
Regarding placement of the URL in the MaRC record, since Prospector does not
load issue level information from the checkin card, the URL for an individual
piece of the title must be sent in the item record if it is desired to have it
show on Prospector. As long as the URL is in the "y" field group tag,
it will be checked by the URL checker.
We need to develop a guidance document on URL placement and usage. To
accomplish this, the next 3 meetings will start early, with a brownbag lunch at
12:30. These will be on March 21, April 13, and May 16.
The regular Feb. Prospector meeting continued after a break for lunch.
Minutes of the Jan. meeting were amended with the additions of two names, Garrison and Gregory, to the attendance list.
Endeavor update
Another timeframe for developing the Endeavor/III interface is in place.
Technical people from Endeavor, III, and the Alliance will talk on March 6 and
over the following 2 weeks to develop a working interface. In April a small
test of Wyoming records will be conducted in the working Prospector system, and
thereafter a test of more complex data. We may need to suppress the relevant
records before the first test load. The many steps involved in this whole
process may mean that the full load of Wyoming records won't be until the end
of the year. It will take about a month for their one million records to load.
NetLibrary update
George has met with the new NetLibrary to discuss the set up of the modified
user based purchase model that was underway last summer; he expects a proposal
from them soon. DPL will do its own profile. Jim Williams will reconvene the
bibliographers committee in March to reconsider the whole issue. George has
asked for bib records, but NetLibrary no longer has a cataloguing staff. OCLC
will provide records, but the conditions are unknown.
DU and CSU completed the load of the owned 7000 NetLibrary titles into their
local catalogs several weeks ago.
GoldRush update
GoldRush is developing the staff module, which should be out in early May. It
will be possible to set view/edit levels of access for individual pieces of the
system, to add and delete holdings and URLs, to see what is owned as a group,
and to see what is in the full text group.
GoldRush should be able to serve as an open URL resolver by the end of 2002. An
open URL resolver is an ISSN table which takes an ISSN from a citation and
looks for it in such places as the public catalogue, Northern Light, or
Expanded Academic.
Enhancements
The Committee reviewed the Circ Committee enhancement requests and amended its request to modify the volume-request function. Circ has requested that volume selection be prompted before the request for patron ID; Cat-Ref added the simultaneous request of multiple volumes.
The Committee would like to resubmit 3 cataloguing enhancements from 2001. #9
from 2001: transfer denied (because of record suppression in Central) searches
from known records to another search key. It was pointed out that a partial
work around is to use l-suppress instead of z-suppress. Transfer to OCLC or ILL
was also discussed, but is not an enhancement request. #2 from 2001: allow the user to view local
checkin records in InnReach. This will allow local libraries to eliminate
individual partner buttons, needed now to get to local holdings. #1 from 2001:
allow flexible Bcode mapping
New enhancement requests: request that
multiple 856 fields in a single record transfer to Central and display in the
institutional record. This should have
been operational in Release 2001, but has never worked correctly. George will
submit this if it is not listed as a known issue; allow a library the option of
forcing patron choice of a pickup site instead of providing a list of sites. As
it stands now, patrons tend simply to accept the first site on the list; allow
indexing of all 022 subfields (a, y, z)
BUT allow overlay only on |a. This may be impossible; George will check.
George will submit the enhancement requests by March 15th to Nancy
Nathanson, the IUG coordinator of the Inn-Reach enhancement process.
The meeting adjourned at 2:30 p.m.