[Sigcr-l] Exhaustivity and specificity of indexing
Marcia J. Bates
mjbates at ucla.edu
Sat Aug 26 22:35:43 EDT 2006
Dear Colleagues,
I first saw the discussion of exhaustivity and specificity of
indexing in Lancaster's 1968 book, Information Retrieval Systems.
It seems to me that the kind of classification system would
play a role in dealing with this question of the exhaustivity and
specificity of classification. Consider a traditional hierarchical
classification. The classification, if properly designed, is assumed
to have classes that are "jointly exhaustive" and "mutually
exclusive." Thus all the intellectual territory to be classified is
supposed to be covered jointly by all the classes in the
classification (jointly exhaustive), and the content of no class
should overlap with that of any other (mutually exclusive). In such
a case, there is one and only one best place in a given
classification to place a given record. There is no way to get
either more exhaustive (one class is the best and only place, so you
won't assign the record to more than one) or more specific (the
document is classified in the one best place--there is no way to make
the class any smaller/narrower.
Since indexing schemes are now sometimes designed to be
faceted, and originally hierarchically organized classifications are
now being at least partially converted to facets--e.g., DDC, there
has been some blurring. One can opt to create narrower classes or
not by utilizing additional facets (DDC, UDC), and records may have
subject terms from several facets assigned, which the online searcher
may combine at will. Under such circumstances, the question of
exhaustivity and specificity is arguably really about INDEXING
ASPECTS of the use of classifications--thus the question is part of
the larger literature of indexing theory. That may be why you don't
find it in the classification literature.
Marcia
--
Marcia J. Bates, Ph.D.
Professor Emerita
Editor (with Mary Niles Maack), Encyclopedia of Library and
Information Sciences
Department of Information Studies
Graduate School of Education and Information Studies
University of California, Los Angeles (UCLA)
Los Angeles, CA 90095-1520 USA
Tel: 310-206-9353
Fax: 310-206-4460
Web: http://www.gseis.ucla.edu/faculty/bates/
More information about the Sigcr-l
mailing list