Re: The Power of Virtuoso Sponger Technology

On 18/10/2009 17:12, "Olaf Hartig" <hartig@informatik.hu-berlin.de> wrote:

> On Sunday 18 October 2009 17:50:58 Hugh Glaser wrote:
>> The SWCL-style approach works pretty well as long as the RDF you want about
>> the URIs is the stuff you get by resolving.
> 
> Right, that's the fundamental assumption. And that's what Linked Data is
> about. ;-)
Yes, but sites say things about other URIs.
In fact, the guidance is to encourage it.
> 
>> It can be much more problematic if the URI is in some site such as (a
>> wrapped) Amazon, saying what is the price of a book identified by a
>> publisher's URI.
> 
> You mean the site didn't mint a URI that identifies the book?
It may well have.
But I don't know what it is.
In the current practice, the site minted a URI for the book, and then
asserted owl:sameAs to other URIs - hopefully including mine, hidden in the
RDF about the site's URI.

In the current SWCL, for example, (I think) if I ask about a dbpedia URI,
the only default place it looks is dbpedia. You can switch on the option to
consult Sindice, and so it may get to some places that Sindice has managed
to find, but that is it. And there are a lot of people making assertions
using dbpedia URIs.

I'm not saying this is the best way to do it (I don't think it is) - I'm
only pointing out that this is what currently happens for the sort of scheme
you describe. It's great as far as it goes, but we should be careful to note
any limitations.

Best
> 
> Olaf
> 
>> There are ways round this, but the technology is not really quite there
>> yet.
>> 
>> Cheers
>> Hugh
>> [...]
> 
>

Received on Sunday, 18 October 2009 16:52:02 UTC