Re: BioRDF: URI Best Practices

On Jul 21, 2006, at 11:38 AM, Xiaoshu Wang wrote:
>
>> notwithstanding, I'd rather know that I am dealing with a
>> non-information resource *before* I touch the network.
>
> I am very puzzled, how can you tell a IR or non-IR given any URI, 
> unless you
> have the knowledge about all URI before hand?  Don't you have to
> de-reference the URI at first hand? '

If you have an ontology, typically the URI is the subject of many 
triples. I assume that some of those triples tell you something about 
what would happen if you dereference the URI.

You are right in the sense that if I receive a naked URI in the email 
I'll have to dereference it to learn something about it. OTOH, this is 
not the case I am thinking about. I am more concerned with URIs that I 
find in a SW context - namely part of a graph - a packet of SW 
information in some message. I expect my ontology to be clear about 
such things as whether a thing is an information resource or not.

>> So my proposal suggests a class that defines ways of transforming
>> the URI you find in a SW document into URLs that get specific types of
> information.
>
> I would also be cautious about that.  This seems to be similar to what 
> the
> web service is doing.  I hope we don't try to reinvent the wheel, 
> especially
> it isn't a small wheel to invent by any means.

Not sure what you mean here. The intention was to provide a mechanism 
for indirection similar to what is desired by the LSID spec, but 
explicitly represented in the same way I represent the rest of my SW 
content, rather than by using another network protocol, like the LSID, 
DNS,  etc.

-Alan

Received on Monday, 24 July 2006 03:51:35 UTC