Aiming at quality and coverage combined: blending physical and virtual union catalogues
| Published date | 01 October 2002 |
| Date | 01 October 2002 |
| DOI | https://doi.org/10.1108/14684520210447895 |
| Pages | 326-334 |
| Author | Janifer Gatenby |
| Subject Matter | Information & knowledge management,Library & information science |
Aiming at quality and
coverage combined:
blending physical and
virtual union
catalogues
Janifer Gatenby
Why library catalogues and union
catalogues?
A common question posed these days is the
relevance of library catalogues when there is so
much information available on the Web that is
directly searchable through Internet search
engines such as Google, AltaVista, Lycos,
Northern Light, etc. To answer this question,
we need to do a comparison. Internet search
engines, for example, have the following
characteristics: lack of precision, retrieval of
poor quality material and poor recall.
Lack of precision
Search engines retrieve too much because they
are retrieving mostly from full text, and to
compensate they rely on relevance ranking.
Library catalogues by contrast index mostly
cataloguing (metadata) that ensure greater
precision. Search engines also retrieve irrelevant
documents. Indexing full text is not precise
because of lack of metadata; e.g. it is not possible
to search a name specifically as an author and
there is usually no controlled subject searching.
At best, there is broad category searching.
Retrieval of poor quality material
Relevance ranking according to number of
times a site has been visited does not
effectively filter out all poor-quality material.
Recall
Relevant material is often buried if it is not
overly descriptive and lacks metadata. Broken
links result in failure to recall relevant
material. Web services and contents of
databases are not touched; therefore the
databases behind the Web pages are not
covered. Resources that are not available on
the Web are not covered. Libraries hold the
key to vast amounts of information, whether
available electronically or not.
The importance of metadata
Library catalogues are based on the tenet that
accurate descriptive and subject cataloguing of
The author
Janifer Gatenby is a Consultant at OCLC/PICA
Information Technology Centre, Leiden, The Netherlands.
Keywords
Libraries, Catalogues, Internet, Data processing
Abstract
This article discusses the deficiencies of search engines
and the importance of metadata before examining three
models of metadata retrieval: distributed; distributed data
with a centralised index; and centralised union catalogue.
In listing the advantages and disadvantages of the
distributed model, the Z39.50 protocol is used as an
example. The OAI harvest protocol is the example of the
second model. Virtual union catalogues are compared
with a real one. A pan-European model is discussed as a
way to combine the best of all three models, with EUCAT
as its base.
Electronic access
The research register for this journal is available at
http://www.emeraldinsight.com/researchregisters
The current issue and full text archive of this journal is
available at
http://www.emeraldinsight.com/1468-4527.htm
Refereed article received 31 May 2002
Approved for publication 19 July 2002
Based on a paper delivered at The European
Library (TEL) ± The Gate to Europe's
Knowledge: Milestone Conference, 29-30 April
2002 Die Deutsche Bibliothek, Frankfurt am
Main, Germany.
326
Online Information Review
Volume 26 .Number 5 .2002 .pp. 326±334
#MCB UP Limited .ISSN 1468-4527
DOI 10.1108/14684520210447895
To continue reading
Request your trial