In article <[log in to unmask]>, Neil
Campling <[log in to unmask]> writes
> I detect a wee bit of arrogance in the view that there is a prescribed or best
>way of doing things. For example, if I was a Yorkshire gardener and wanted to
>find out about goke rot (goke is only one of several possible spellings
>depending upon the pronunciation), you tell me what meta-data set would have the
>nouce to direct me to the most appropriate authority ?
Now we're getting back to first principles. The idea of an information
retrieval thesaurus or classification scheme is that you are not
organising _words_, you are organising _concepts_.
I don't know what goke rot is, but presume that it is some form of rot
that affects plants. It is a concept, and anyone who knows about such
things would recognise it when shown the rotten carrot or whatever.
Different people might call it by different names, but that doesn't
change what it is : "A rotten carrot by any other name would smell as
foul".
A classification scheme or thesaurus builds relationships between
concepts, and the structure is therefore in principle independent of
language. We have to label each concept to make it easy to refer to, and
the easiest way to do this is to choose a concise word or phrase that
will be generally understood to identify that concept. The choice of
such a term is fairly arbitrary and calling it a "preferred term" does
not imply that it is better than any other, just that it has been chosen
as the term most likely to be identified with the concept by the
majority of users of the indexing system. If there is any ambiguity, the
definition of the concept, as used in the indexing system, must be
spelled out in a scope note.
Other possible terms which might be sought for the same concept must
also be included in the indexing vocabulary, with USE references to the
one that has been chosen. In many classification schemes the concept
label is an alphanumeric symbol and _all_ the terms that people would
actually look for have to be linked to this in an index.
If you are looking for "goke rot" in a system that uses a good
controlled vocabulary, it might be included as an entry or indexing term
which will take you directly to any relevant information. If not, you
can scan the concepts that are listed under the most specific terms that
are there, such as "plant diseases" and may find it under another name -
the scope note will help you decide if it is the same concept. Thus the
thesaurus helps you to find information even though it does not contain
the term you first thought of.
Having said all that, I would not deny that there is a place for
uncontrolled free text searching - the two techniques complement each
other - but that is a discussion for another day.
Leonard Will
--
Willpower Information (Partners: Dr Leonard D Will, Sheena E Will)
Information Management Consultants Tel: +44 (0)20 8372 0092
27 Calshot Way, Enfield, Middlesex EN2 7BQ, UK. Fax: +44 (0)20 8372 0094
[log in to unmask] [log in to unmask]
---------------- <URL:http://www.willpowerinfo.co.uk/> -----------------
|