Re: Supporting arrays of concepts

In message <6.0.0.22.2.20040510175606.01bf9ac8@pop.skynet.be> on Mon, 10
May 2004, Ron Davies <ron@rondavies.be> wrote
>Leonard,
>
>I won't argue with you, except to say that whether these are "node labels"
>or "guide terms", we can't ignore them. These are not, in the Getty way of
>thinking, concepts, and they function, in terms of the hierarchy, in ways
>very similar to node labels.

Ron -

Only in the respect that they are terms that Getty says should not be
used for indexing. They do have BT/NT and RT relationships, which true
node labels do not. In the example you quoted,

>>furnishings
>>      <furnishings by form and function>
>>           <coverings and hangings>
>>                <coverings and hangings by general type>
>>                     coverings
>>                     hangings

<furnishings by form and function> and <coverings and hangings by
general type> are of course proper node labels.

Further down the same hierarchy, there is a real example of the problem
you raise, with one node label immediately under another:

<coverings and hangings by specific type>
   <coverings and hangings by form>
       blankets (coverings)
       rugs
       throws (coverings)
   <coverings and hangings by function>
      dust covers
      shrouds
      sleeping bags

(I have omitted some entries which obscure the issue).

In this case I think that the node label <coverings and hangings by
specific type> is redundant and could be omitted.

>The question remains as to how this situation is to be described in RDF?

If you want to encode AAT as it stands, then yes, you have to find a
structure to cope with this.

Should we, though, be creating a structure that allows encoding of
inconsistent and illogical structures just because they occur in some
existing thesauri? Library of Congress Subject Headings has some
thesaural elements, but nobody would argue that it conforms to thesaurus
standards. Should we be able to encode it in RDF?

Is it unrealistically idealistic to develop an encoding system based on
standards, or unrealistically complex to develop a system that takes
account of all the idiosyncrasies that may be found in existing
thesauri?
(I don't know the answer - no doubt we have to find a compromise!)

>Incidentally, I didn't go looking for this example, it popped up in a quick
>five-minute browse, which suggests that there must be quite a few more in
>the AAT (which I don't know at all).

Yes, there are lots. I suggest that we should encode elements of this
thesaurus according to what they are, rather than what the publishers
say they are, i.e.

*       Treat real node labels of the form <xxx by yyy> as node labels,
        and allow for one of them occurring directly under another.

*       Treat other terms in angle brackets as thesaurus terms, but add
        a note "Do not use for indexing". We have in any case to allow
        for editorial and usage notes of this kind as well as for scope
        notes.

Leonard

-- 
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)870 051 7276
L.Will@Willpowerinfo.co.uk               Sheena.Will@Willpowerinfo.co.uk
---------------- <URL:http://www.willpowerinfo.co.uk/> -----------------

Received on Monday, 10 May 2004 14:37:18 UTC