W3C home > Mailing lists > Public > public-rdf-dawg-comments@w3.org > March 2009

RDF Syntax for SPARQL

From: Holger Knublauch <holger@topquadrant.com>
Date: Thu, 5 Mar 2009 11:19:15 -0800
Message-Id: <CFD49C59-373A-49C2-8A4C-29CA3213FA31@topquadrant.com>
To: public-rdf-dawg-comments@w3.org
Dear SPARQL Working Group,

I would like to draw the group's attention to an RDF Syntax [1] that I  
have developed as part of the SPIN [2] language family. The SPIN RDF  
Syntax provides an RDF Schema for representing all SPARQL (1) elements  
such as queries, triple matches and filters, plus a couple of Jena/ARQ  
extensions such as assignments and aggregations. We have a reference  
implementation of this syntax and will make it open source shortly.  
TopQuadrant (W3C member) would welcome if the working group could  
discuss making this (or a similar) RDF notation a standard. We are  
aware of the limited time of this working group, but we regard this  
spec as already sufficiently mature, as we have it in routine use in  
our TopBraid Suite tool family, and many of our users have provided  
positive feedback already. SPIN will be officially announced with  
TopBraid 3.0 release scheduled for April 2009.

The SPIN RDF Syntax is similar to languages like SWRL and OWL that  
represent various language constructs in terms of RDF blank node  
trees. See the linked documents for examples. There are also numerous  
examples of SPIN in my blog [3].

There are various reasons why I believe such an RDF schema will be  
useful. In particular it makes it possible to consistently store  
SPARQL queries together with RDF/OWL models. As a result of this,  
SPARQL can be leveraged for completely new use cases such as SPARQL as  
a rule language, SPARQL as a constraint checking language, and  
reusable SPARQL templates as demonstrated by the SPIN Inferencing and  
Modeling Vocabulary. Storing SPARQL queries in RDF avoids  
complications of namespace prefixes (the namespaces of the surrounding  
RDF document can be used). Furthermore, having SPARQL queries as  
proper RDF resources makes it easier for tools to find references to  
URIs, for example to detect which parts of a SPARQL query need to be  
modified when a resource gets renamed. Finally, SPARQL queries can be  
provided with a URI and thus shared on the Semantic Web. Needless to  
say there is a direct rendering from the RDF syntax to SPARQL textual  
syntax for editing.

I have recently heard that there are suggestions for an XML syntax for  
SPARQL (queries). Using the RDF/XML serialization, SPIN RDF queries  
can be rendered into XML as well, but also into (quite readable) N3.

I would be happy to discuss this technology in detail with the group.

Regards,
Holger

[1] http://spinrdf.org/sp.html
[2] http://spinrdf.org
[3] http://composing-the-semantic-web.blogspot.com
Received on Thursday, 5 March 2009 19:20:00 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Thursday, 5 March 2009 19:20:00 GMT