For discussion, on behalf of a non member...Link to Holdings & ISSN/ISBN

[Don Hamilton]

The following was sent to Silverplatter's spin-l list by a compatriot of
mine, who is trying to address the issue of inconsistent issn/isbn data and
the problems with hook to holdings using those datum. He has received few
responses, so I am posting his comments here too (sorry for the duplication)
in the hopes of broadening the search for suggested approaches to an
automated solution.

Comments?

Don Hamilton(dhamilto at mach2.wlu.ca)
Information Technology & Librarian by Osmosis
Wilfrid Laurier University/Tri University Group
519-884-0710-x3336, fax 519-884-8023D


WebLink is a truly great feature!

Since its functionality is totally dependent on exact matching ISSNs/ISBNs,
there are - not unexpectedly - (many) instances where a user does not get a
match even though a record for the title exists.  The reasons for this may
be as simple as lack of an ISSN/ISBN in the cataloguing record, the correct
ISSN/ISBN in a catalogue record but in the wrong format (sometimes
including extra data), and simply a wrong/old ISSN/ISBN.  The big,
underlying reason is undoubtedly that the cataloguing of ISSN/ISBN data has
been subject to varying practices at over time and that these also differ
between libraries.  The increasing importance of ISSN/ISBN data as the
'hook' for holdings checking, not only with ERL but other services as well,
certainly brings what has been a fairly hidden problem with cataloguing
data integrity to the forefront.  There is an increasing need and pressure
to verify/update/correct at least ISSN data; (the feasibility of doing this
for ISBNs would appear to be less than practical and the benefits are
simply not as great in comparison with ISSNs, at least on a per title
basis).

The question is:  how to verify/update/add ISSN data in a library catalogue
in the most effective manner?

My hope is, since this problem cannot possibly be unique to this ERL site
(TUG - U. of Guelph, U. of Waterloo, Wilfrid Laurier U.), that this message
will bring some suggestions and/or descriptions of what can/has been done
elsewhere to meet the need for accurate ISSN data in cataloguing records,
or simply whether it is possible to automate the process of ISSN
verification (and updating) as much as possible.

Handling the process of ISSN verification/updating entirely in a manual
fashion would require significant and likely unavailable resources in most
cases, particularly large academic and research libraries.  At the very
best, such a process would take a very long time to complete ...

Any reduction of the amount of manual, item by item checking of cataloguing
records and corresponding copies would increase the feasibility (and
likelihood) of ISSN verification/updating.  While it cannot reasonably be
expected that an automated approach would be 100% dependable and accurate,
it could serve to at least eliminate that portion of records from the total
list of serial holdings where the ISSN data is correct, leaving a (not
necessarily small) residue that requires further attention, a portion of
which could be further processed by additional automated
procedures/algorithms (perhaps with title matching and human verification),
etc. ...

These are just some thoughts ...  I would very much appreciate any
reactions/responses and/or suggestions/pointers, etc.  In a large or small
way, all libraries will be faced with this problem AND how to deal with it
sooner or later ...

Thank you!


==============================================================
Ralph Daehn                 E-Mail:  rdaehn at UoGuelph.CA
Library, Univ. of Guelph    Voice:  (519) 824-4120 Ext. 3603
Guelph, Ontario  N1G 2W1    http://www.lib.uoguelph.ca/~rdaehn
==============================================================