Re: What if an URI also is a URL

Am Freitag, den 14.09.2007, 20:06 +0200 schrieb Richard Cyganiak:
> On 14 Sep 2007, at 00:01, Oskar Welzl wrote:
> > The trouble is finding the right ones for RDF you write yourself.
> > "Right" in this context means there's a chance of other's using the  
> > same
> > URI for the same thing, which *should* be easier for things 'on the  
> > web'
> > than in general. Or so I thought before...
> 
> Hm, I don't see the trouble.
> Want to find a URI for something?
> If you know that it is an existing web resource, use its URI.
> 

So whats the URI for flickr, the service that allows you to upload
photos? <http://www.flickr.com>? Or rather <http://flickr.com>? Oh no,
wait a minute, we agreed before we shouldnt use the URI of a single
information resource when we refer to a whole site/service. So -
<http://my.domain.net/id/websites#flickr>?

It's *not* easy and I'm hating it. ;)

> You should never be afraid of minting new URIs, even if there might  
> be an existing one out there already. If you learn about a good  
> existing URI later, then you can still declare it owl:sameAs. Not  
> knowing a good existing URI for something should never stop you from  
> publishing RDF data about that thing.

Right. Its hardly a problem if we use different URIs for the same thing.
It becomes nasty, though, if we use the same URI for different things -
as it happens when I say something is a foaf:homepage (=a
foaf:document), while somebody else claims the same thing is a
sioc:Forum and, as such, contains sioc:Posts (that are not on my
weblog's front page and never were).
Again, from an RDF-consumers' point of view, none of this matters; the
same URI can also be used to describe the weather in Bern yet again.
It's the authors who have to take care. That's why I said it's more
difficult to choose the right URIs when authoring than to read/interpret
them in any given document.

> That being said, there are lots of great resources for finding  
> existing SemWeb-friendly URIs. You can use the SemWeb search engines,  
> e.g. SWSE, Swoogle, Uriqr. Lots of good stuff can be found around the  
> Linking Open Data project. If it exists in Wikipedia, then it has a  
> URI in DBpedia.

Those are great, yes. I started falling in love with dbpedia because
its so "real life".

BTW: I highly appreciate your feedback here.

Oskar

Received on Tuesday, 18 September 2007 20:11:19 UTC