RE: Concept spaces and Namespaces RE: Glossary of terms ...

Isn't a space the place where you can find something laid out, such as a
scheme? And doesn't a scheme have an existence (in the abstract of
course) independent of the place where it is laid out? (Even if one of
those places is designated as more authoritative than the others.)

And yes, some people do already use the term "concept space" with a
slightly different meaning - something I vaguely associate with using
vector analysis to place concepts in an n-dimensional concept space,
only I don't have the energy to go and check. Sorry!
Stella

*****************************************************
Stella Dextre Clarke
Information Consultant
Luke House, West Hendred, Wantage, Oxon, OX12 8RR, UK
Tel: 01235-833-298
Fax: 01235-863-298
SDClarke@LukeHouse.demon.co.uk
*****************************************************



-----Original Message-----
From: public-esw-thes-request@w3.org
[mailto:public-esw-thes-request@w3.org] On Behalf Of Miles, AJ
(Alistair)
Sent: 18 January 2005 17:09
To: 'public-esw-thes@w3.org'
Subject: RE: Concept spaces and Namespaces RE: Glossary of terms ...



Regarding the question of the relationship between concept spaces and
namespaces, +1 on what Chaals said.

Considering the idea of moving from talking about 'concept schemes' to
'concept spaces' for a moment:

 - We could work from a definition such as 'a concept space is a set of
concepts, and may include a set of relationships between those
concepts.' As far as I know, this is consistent current usage of
'concept space'.

 - However, 'concept space' is currently mostly used (as far as I could
find) to describe a network of terms (sic) with collocation metrics
automatically generated by a computer algorithm from a document set.  

 - The 'skos:inScheme' property would need to be replaced too, with
something like 'skos:inSpace' or just 'skos:space' (I hate it when I
can't think up good property names).  

 - Replacing the class 'skos:ConceptScheme' with 'skos:ConceptSpace' and
replacing 'skos:inScheme' with something else *is feasible* I think
(i.e. could be done without causing too much disruption), as the
semantics are essentially the same, meaning that the old and new
property sets could happily cohabit for as long as it takes, with
owl:equivalent... statements linking them.

Do we want to do this?  How much do we want to be able to talk about
'concept spaces'?

Incidentally, I was also pondering changing 'skos:Collection' to
'skos:ConceptGroup' or something like that, (what about
'skos:ConceptSubSpace'? just kidding :) to avoid overlap of 'collection'
with RDF usage and with the library/museum notion of a 'collection'.
Anyone have any opinions on this?

Cheers,

Al. 

---
Alistair Miles
Research Associate
CCLRC - Rutherford Appleton Laboratory
Building R1 Room 1.60
Fermi Avenue
Chilton
Didcot
Oxfordshire OX11 0QX
United Kingdom
Email:        a.j.miles@rl.ac.uk
Tel: +44 (0)1235 445440



> -----Original Message-----
> From: public-esw-thes-request@w3.org 
> [mailto:public-esw-thes-request@w3.org]On Behalf Of Charles 
> McCathieNevile
> Sent: 18 January 2005 16:15
> To: Mark van Assem
> Cc: Bernard Vatant; 'Thomas Baker'; Miles, AJ (Alistair); 
> public-esw-thes@w3.org
> Subject: Re: Concept spaces and Namespaces RE: Glossary of terms ...
> 
> 
> 
> On Tue, 18 Jan 2005, Mark van Assem wrote:
> 
> >Salut from me too,
> >
> >> I think that it is not a good idea for the SKOS
> specification to try and
> >> equate a namespace with a set of concepts described as a
> "scheme". If I want
> >
> >If I understand namespaces correctly, they are only a way to provide 
> >unique names. If so, a particular namespace is not a
> "coherent" set of
> >classes and properties. (Although this usually is the case,
> a namespace
> >does not imply this.) The SKOS "concept scheme/space" does have this 
> >stronger meaning (if I interpret the spec correctly).
> 
> Right. In particular, if I am a user, I might want to create a SKOS 
> collection of concepts/terms, which is mostly built from existing ones

> that are identified by URIs that happen to have two or three
> "namespaces". I don't
> see any reason to make a new namespace to duplicate these in 
> a single place,
> beyond aesthetic beauty for people who are perverse enoughto 
> want to read
> the underlying code rather than just get on with their real 
> work of using the
> collection.
> 
> >>>name *inside the concept space* before being ported to the
> RDF format. So an obvious
> >>>migration practice will certainly be to use a single RDF
> namespace to somehow represent
> >>>the concept space. I don't know if that should be
> recommended by the specification, or
> >
> >Probably it is a good practice, but if the above argument holds, the 
> >concept scheme and inScheme property are still needed to provide the 
> >stronger meaning of a coherent set.
> 
> I think the argument that the SKOS stuff is necessary is
> strong. Further,
> there is no good reason I can see why I should not be able to 
> use the same
> namespace prefix for my SKOS collection, my ontology for 
> discussing medieval
> cookery, and a couple of extension terms I might write to 
> refine Inkel's
> vocabulary dealing with the languages people speak. The URI 
> is just a string
> used to provide a unique name, right?
> 
> cheers
> 
> Chaals
>

Received on Tuesday, 18 January 2005 19:02:52 UTC