Diane I. Hillmann <[log in to unmask]> wrote:
>> I see DC as a foundation upon which much can be built. We should not
>> discourage people to use DC, even when it goes beyond the simple uses that
>> we intend it for. Instead, we should encourage such development, as it will
>> strengthen the semantic meaning of documents.
> Not all of us agree with this. See:
> http://www.dlib.org/dlib/january01/lagoze/01lagoze.html (with which I
> heartily agree, by the way).
It seems that the problems raised in this document stem from a confusion of
has and is. In RDF, the issues are resolved quite cleanly. Let's take the
Cornell example:
Item 1:
creator = Gary Cornell
Item 2:
creator = Allison Lurie
creator.organization = Cornell University
Using this system, it is unclear whether creator.organization specifies a
type of creator (is) or more information about the creator (has).
In RDF, we'd write it as:
<#1> b:author "Gary Cornell"
b:author rdfs:subClassOf dc:creator
<#2> dc:creator <#INTNODE>
<#INTNODE> rdf:value "Allison Lurie"
<#INTNODE> sq:organization "Cornell University"
Using such a system, we all win. Those who want simple metadata can easily
retrieve it through rdf:value, those who need something more complex have it
available to them.
--
[ Aaron Swartz | [log in to unmask] | http://www.aaronsw.com ]
|