- From: Pasqualino Assini <assini@kamus.it>
- Date: Thu, 25 Apr 2002 12:35:21 +0100
- To: <www-rdf-interest@w3.org>
RDF descriptions are typed and types/classes is what you should use to specify what you are talking about. It what you want to talk about is "IBM the company" you would create a class Company, specifying in its human readable description (the rdfs:comment) what is its semantic and then write something like: <Company rdf:about="http://www.ibm.com/company"> ... whatever you want to say about IBM the company </Company> If you want to talk about "IBM the share" you would define a "Share" class and say: <Share rdf:about="http://www.ibm.com/share"> ... whatever I want to say about IBM the share </Share> The URLs are there just to provide an unique IDs for these distinct concepts. A URL naturally has also a content (that for HTTP would be what you recieve from a GET command) and it is both very simple and very useful, though entirely optional, to make sure that the content of the resource is an RDF description. This would be the "official" description of the entity as provided by its owners. So if you access http://www.ibm.com/company you would get what IBM wants to tell you about itself. Naturally people can use different URLs and different classes to refer to the same real-world concept (such as "IBM the company"). This cannot be avoided but naturally is possible to add an additional layer of metadata to establish equivalences across different ontologies and naming systems. titto ------------------------- Pasqualino "Titto" Assini - Nesstar Ltd John Tabor Building - University of Essex Colchester, Essex - CO4 3SQ - United Kingdom email: titto@nesstar.com personal email: titto@kamus.it
Received on Thursday, 25 April 2002 10:02:37 UTC