Message-Id: <199708101747.MAA11726@library.wustl.edu> Date: Sun, 10 Aug 1997 12:51:55 CDT From: Gerry McKiernan <mailto:JL.GJM@ISUMVS.IASTATE.EDU> Subject: Diamonds-in-the-Rough(sm): Semantic Nets for Subject Thesauri To: Multiple recipients of list WEBCAT-L <mailto:WEBCAT-L@WUVMD.WUSTL.EDU>
Diamonds-in-the-Rough(sm):
Semantic Nets for Subject Thesauri
For my review of 'neo-conventional' thesauri [never-ending] I
I am interested in efforts that have applied Semantic Networks
or Conceptual Graphs to faceted (or non-faceted) _subject_ thesauri.
A Semantic Net may be described as a graphic structure that
represents the semantic associations of words and/or
concepts within a textual corpus [Over-simplified].
I have been inspired to submit this posting from my
reading of a fascinating study by Sedelow and Sedelow
of the University of Arkansas at Little Rock (USA) reported in
their paper "Thesauri and concept-lattice semantic nets"
[Advances in Knowledge Organization 4(1994): 350-357]
In their paper, S & S describe a prototype in which they
have applied formal concept lattices to construct a
'rigorous and empirically accurate semantic net(s)'.
In their prototype they the concept lattice approach to
for 'representing the implicit, or internal structure, of
_Roget's International Thesaurus_ (3rd edition) [1962]
[This is contrasted with the "explicit hierarchical structure
inasmuch as empirical investigation has shown that the
explicit upper levels in the hierarchy are in many ways open
to question ...]
This insight for the Roget's Thesaurus has the potential
of identifying, creating _and_ graphically displaying the
cross-structural relationships that I believe are not offered
in many conventional _subject_ thesauri [Of course, here
I am assuming that the 'neo-relationships' uncovered by
the creation of a semantic net and an associated graphic
or other visual display, will indeed be more effective that
the current conventional structures and forms of display]
[This of course Remains to Be Seen [get it, 'Seen' - 'Display']
[:->]
I am also aware of [have not read] the work of Roya Rada et al.
"Retrieval Hierarchies in Hypertext" _Information Processing
and Management_ (v29 n3 p359-71 May-June 1993)
and of course, the highly-innovative Unified Medical Language
System (UMLS) developed at the NLM.
As always, and leads, citations,suggestions, reactions,
criticism, critiques, opinions, gems [Sayings or Precious
Stones] are welcome!
<strong> Thanks </strong>
Gerry McKiernan
Curator, CyberStacks(sm)
Iowa State University
Ames IA 50011
mailto:gerrymck@iastate.edu
http://www.public.iastate.edu/~CYBERSTACKS/
"The Best Way to Predict the Future is To Invent It!"
Attributed to Peter Drucker