Re: RE : aboutness and broader

On Fri, 6 Jun 2008 at 11:56:41, Antoine Isaac <Antoine.Isaac@KB.nl> 
wrote
>I agree that the examples given by Jonathan are a bit unclear, but I 
>think the concern is genuine, and partly share it.

I'm sorry, but I don't understand what the cause for concern is here.

Alistair said:

><D> dc:subject <A>.
><A> skos:broader <B>.
><B> skos:broader <C>.
>
>entails
>
><D> dc:subject <B>, <C>.
>
>Notice also that this behaviour only depends on the property chain 
>axiom. It does not require that skos:broader be transitive.

and this seems perfectly reasonable.

>In library cases, having a document's subject index automatically added 
>with the parent concepts of the ones that were chosen by the indexer 
>would violate the indexing specificity principle.

In any case (this is not specific to library applications) this does not 
imply that the document has to be indexed with all the broader concepts. 
By linking the most specific concept(s) in each relevant part of the 
subject hierarchy to the document, the indexing specificity principle is 
maintained. The retrieval system should then be able to refer to the 
thesaurus to find broader concepts if necessary, or to "explode" the 
search to include narrower concepts if the search has been expressed in 
less specific terms.

Some retrieval systems do not have the capability to use a thesaurus in 
this way, and then it may be necessary to link complete hierarchical 
chains of terms to a document, but that is just to overcome a limitation 
of the system.

>Notice that this problem is a bit similar to the choice between making 
>skos:broader transitive and introducing a new 'transitive version of 
>it' to avoid messing with the original intended meaning of concepts.

The SKOS documentation on this is somewhat confusing. It  says, at 
example 36, "skos:broader is a sub-property of skos:broaderTransitive" 
and  "...a sub-property of a transitive property is not necessarily 
transitive". Perhaps I don't understand the inheritance of 
characteristics in SKOS properties, but it sounds anomalous to say that 
a general type of relationship such as "broader" should be a 
sub-property of a more restricted type of relationship such as 
"broaderTransitive". I realise that this issue is still under 
discussion, so perhaps it will be sorted out eventually.

>If we enforce for knowledge bases the kind of rule that Alstair 
>proposes, then it's more difficult to make the distinction between 
>asserted subjects and inherited ones, hampering all kind of interesting 
>options wrt to query expansions, as explored in Uni-Glamorgan's work 
>[1] and Alistair's thesis [2].

>Maybe a special type of property (in this case, different from 
>dc:subject) would be appropriate fo allowing these subject inferences, 
>which are otherwise very useful in a number of cases of course (e.g. 
>classical faceted browsing). Or maybe just mention that the rule is 
>just an indication on how to add interesting (but not mandatory) 
>functionality to a knowledge base...

I'm not clear why you want to make this distinction or what different 
property you see the need for. If a document is about mammals, it is 
about animals. How are query expansions hampered by a lack of 
distinction between asserted and inherited subjects?

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 Friday, 6 June 2008 11:07:54 UTC