W3C home > Mailing lists > Public > www-rdf-interest@w3.org > May 2003

RE: URIQA!

From: <Patrick.Stickler@nokia.com>
Date: Thu, 22 May 2003 10:00:05 +0300
Message-ID: <A03E60B17132A84F9B4BB5EEDE57957B01B90E1F@trebe006.europe.nokia.com>
To: <leo@ist.org>, <piotr@ideanest.com>, <www-rdf-interest@w3.org>



> -----Original Message-----
> From: ext Leo Sauermann [mailto:leo@ist.org]
> Sent: 22 May, 2003 03:24
> To: 'Piotr Kaminski'; www-rdf-interest@w3.org; Stickler Patrick
> (NMP/Tampere)
> Subject: RE: URIQA!
> 
> 
> looks like a nice implementation !
> (tomcat / jena2)

Thanks. I'm a fan of both.

> > Why restrict all replies to a concise bounded description?  
> > If an agent is
> > interested in information "around" a resource, right now it 
> > would have to
> > issue a (potentially large) number of individual requests.  
> Adding an
> > optional "expansion depth" parameter to the GET query would 
> > provide a simple
> > way to grab information in bigger chunks.  
> 
> hm, I know the "depth" value from 
> http://www.musicbrainz.org/MM/index.html
> 
> they have many APIs that are based on it:
> http://www.musicbrainz.org/products/client/download.html
> 
> stick to the words of the wise man:
> "if you enter a depth of 10, you will get all data of the 
> world. please
> wait..."
> serious: you could connect every person in 6 steps, so with data it
> could be the same.

Exactly. The URIQA model is deliberately constrained to knowledge about
a specific resource. Particular implementations and services can (and
surely will) provide additional means for arbitrary queries of their
knowledge bases.

URIQA tries to capture that core, essential behavior that one can hope
to expect from every Semantic Web enabled server, and to that end, less
is more.

A SW agent can then interact in a consistent manner with any arbitrary
URIQA enlightened server, irrespective of the additional services and
capabilities each particular server might offer. It also allows the
SW agent to decide itself just how "deep" it will probe the SW for
the knowledge it needs.

Constraining queries to concise bounded descriptions is a key to
controlling infoglut for SW agents, where otherwise, a simple query
can result in (potentially) thousands or tens of thousands or even
more triples when all it ever wanted was a rdfs:label or an rdf:type.

> And 2:
> 
> I assume you know Sesame and its web access model, if not:
> http://sesame.aidministrator.nl/publications/users/ch07s02.html
> they used simple query fields. this has advantages and problems,
> so I ask a "dumb" question to hear your ideas:
> 
> What do you think of, who should use the Sesame model and who 
> the URIQA
> model ?

Those who wish to submit arbitrary queries to a knowledge server will
want to use solutions such as Sesame -- yet Sesame can also (hint hint ;-)
provide a URIQA interface for providing concise bounded descriptions
of resources.

You'll note that the demo implementation of URIQA includes an advanced
search page (still a work in progress) which will in fact provide for
arbitrary queries in several query languages. But that is a value added
extension to the URIQA model, and specific to that particular implementation.

One could easily use Sesame as the foundation for a URIQA enlightened server,
while still providing all of the additional functionality of Sesame.

> (whoa, this is groovy feeling, seeing a standard in its baby stage)

From your lips to the industry's ears ;-)

Cheers,

Patrick


> greetings
> Leo Sauermann
> Vienna, Austria
> 
> 
Received on Thursday, 22 May 2003 03:00:09 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Monday, 7 December 2009 10:51:59 GMT