RE: naming labelled instances

Hi Kevin,

Sure, it makes sense to use a base namespace in a schema, but its a little
harder to know what to do in instance data. What I am trying to understand
is what is best practice for creating controlled vocabularies. Here rather
than having a set of controlled vocabularies ahead of data ingest e.g. for
validation, we are creating the vocabularies during ingest. So we want
namespace collisions to occur if we are referring to the same thing, but if
namespace collisions occur and they are not referring to the same thing we
have a problem. 

For example in the faceted browse tool for schemas John G put together, one
of the facets was namespace and another was fragment. So this tool (with a
bit of work) would let you browse all the topics or the geographies on a
property by property basis. However this is done by virtue of the URI, which
feels like cheating. 

Another way to do this would be to create two classes, say TopicValues and
GeographicValues, and then say each property value is a class and subclass
it to the appropriate class. This would declare the controlled vocabulary in
a RDFS way. The only problem is in RDFS there is no way to distinguish
between classes that have properties e.g. are used for instances and classes
that are property values. 

Dr Mark H. Butler
Research Scientist                HP Labs Bristol
mark-h_butler@hp.com
Internet: http://www-uk.hpl.hp.com/people/marbut/


> -----Original Message-----
> From: Kevin Smathers [mailto:kevin.smathers@hp.com]
> Sent: 21 October 2003 18:10
> To: Butler, Mark
> Cc: 'www-rdf-dspace@w3.org'
> Subject: Re: naming labelled instances
> 
> 
> As a rule, what I see in other documents is an entity defined to hold 
> the base name of each namespace, and everything else is based on that 
> entity.  In IMS for example I find this:
> 
> <!DOCTYPE rdf:RDF [
>          <!ENTITY rdf      
> 'http://www.w3.org/1999/02/22-rdf-syntax-ns#'>
>          <!ENTITY rdfs     'http://www.w3.org/2000/01/rdf-schema#'>
>          <!ENTITY dc       'http://purl.org/dc/elements/1.1/'>
>          <!ENTITY dcterms  'http://purl.org/dc/terms/'>
>          <!ENTITY lom      'http://ltsc.ieee.org/2002/09/lom-base#'>
>          <!ENTITY lom-edu  
> 'http://ltsc.ieee.org/2002/09/lom-educational#'>
> ]>
> 
> The only difference from your example is that the author IMS tends to 
> prefer dash over underbar, and prefixes each name with the 
> name of the 
> overall schema.  If you were to follow the same naming 'id' would be 
> 'vra-id', source would be 'vra-source', etc.
> 
> Cheers,
> -kls
> 
> Butler, Mark wrote:
> 
> >Hi team,
> >
> >At the teleconference last week we discussed that sometimes it is
> >advantageous to name common property values with URIs, then 
> use labels to
> >give textual versions of those property values. I've added 
> this to the
> >Artstor stylesheet, and I've been using Brownsauce to 
> investigate the data
> >as Eric suggested. However does anyone have any 
> recommendations on how to
> >name these URIs? At present I am using a schema namespace of
> >
> >http://www.dspace.org/simile/schemas/artstor-06oct#
> >
> >and various property value namespaces e.g.
> >
> >http://www.dspace.org/simile/metadata/artstor/id#
> >http://www.dspace.org/simile/metadata/artstor/source#
> >http://www.dspace.org/simile/metadata/artstor/image_id#
> >http://www.dspace.org/simile/metadata/artstor/object_id#
> >http://www.dspace.org/simile/metadata/artstor/mediafile#
> >http://www.dspace.org/simile/metadata/artstor/resolution#
> >http://www.dspace.org/simile/metadata/artstor/mediafileformat#
> >http://www.dspace.org/simile/metadata/artstor/collection#
> >http://www.dspace.org/simile/metadata/artstor/creator#
> >http://www.dspace.org/simile/metadata/artstor/material#
> >http://www.dspace.org/simile/metadata/artstor/recordType#
> >http://www.dspace.org/simile/metadata/artstor/site#
> >http://www.dspace.org/simile/metadata/artstor/geographic#
> >http://www.dspace.org/simile/metadata/artstor/format#
> >http://www.dspace.org/simile/metadata/artstor/topic#
> >
> >with the property value coming after the hash.
> >
> >Now ignoring the question of the correct stem for the 
> namespace, which we
> >also discussed at the telecon, I'm not sure if I have chosen 
> my fragements
> >and namespaces correctly. Is it good practice to do it this 
> way? Would it be
> >better to have the property value type after the hash and 
> have a single
> >namespace? Alternatively is the schema we use for 
> namespacing a irrelevance,
> >and as long as we don't have namespace collision is everything ok?
> >
> >One "nice to have" would be to browse by topics e.g. list 
> all the topics in
> >this dataset. However this would seem to require subclass 
> relations, right?
> >Or is there another way of doing this?
> >
> >kind regards
> >
> >Dr Mark H. Butler
> >Research Scientist                HP Labs Bristol
> >mark-h_butler@hp.com
> >Internet: http://www-uk.hpl.hp.com/people/marbut/
> >  
> >
> 
> 
> -- 
> ========================================================
>    Kevin Smathers                kevin.smathers@hp.com    
>    Hewlett-Packard               kevin@ank.com            
>    Palo Alto Research Lab                                 
>    1501 Page Mill Rd.            650-857-4477 work        
>    M/S 1135                      650-852-8186 fax         
>    Palo Alto, CA 94304           510-247-1031 home        
> ========================================================
> use "Standard::Disclaimer";
> carp("This message was printed on 100% recycled bits.");
> 
>

Received on Tuesday, 21 October 2003 13:36:51 UTC