- From: Seaborne, Andy <andy.seaborne@hp.com>
- Date: Wed, 7 Jun 2006 09:26:33 +0100
- To: "Chris Bizer" <chris@bizer.de>, <connolly@w3.org>, <danbri@danbri.org>, <danny.ayers@gmail.com>, "Richard Cyganiak" <richard@cyganiak.de>
- Cc: <www-archive@w3.org>, <Lehmann.Christian@gmx.net>, <eric@w3.org>
-------- Original Message -------- > From: Chris Bizer <mailto:chris@bizer.de> > Date: 6 June 2006 11:26 > > Hi all, > > I also cced Andy Seaborne, Eric Prud'hommeaux, Richard Cyganiak and > Christian Lehmann. To them, this mail is in reply to: > http://dig.csail.mit.edu/breadcrumbs/node/140 > and > http://lists.w3.org/Archives/Public/www-archive/2006Jun/0003.html > > We are of course very interested in joining efforts on database to RDF > mapping. So a sceduled chat or a BOF at ESWC are both good ideas. A scheduled chat would be good - I won't be at ESWC. > > We plan to do the following stuff for D2RQ and D2R server in the next > months: > - implement a smarter SPARQL-to-SQL rewriting algorithm, which will > speed up queries. Richard's "A relational algebra for SPARQL" was the > first step into this direction. > - do some fine tuning on the language and introduce some language > constructs that make it easier to write mappings or make the language > more flexible. > Did anybody hear of use cases where something couldn't be mapped using > D2RQ? > - improve Richard's tool for the automatic generation of mappings. > - write a nice setup routine which will make it easyer to get the > server running. > - publish additional mappings for commonly used wikis, blogs and CMSs > in order to speed up the creation of public SPARQL endpoints. Damian has just released: http://jena.sourceforge.net/SquirrelRDF/ which enables non-RDF data to be queried in SPARQL. Backends for SQL and LDAP. LDAP is a great source of info for us industry types. Richards incorporated the SQL metadata extractor into D2RServer, if I understand correctly. I wonder if the D2RQ mapping can be separated from the SQL-access so that adapters can be added for other sources. Or whether it is better to take a rules-based approach and map "low level" RDF (a direct translation of the base source) into "high level" RDF (better data modeling). Is the mapping always the same for all applications? > > So if anybody knows about related work or wants to contribute ideas to > any of the topics, everything is welcome. > > Christian Lehmann (University Leipzig) is working on a D2RQ > imlementation in PHP. It would be great to be able to use D2RQ in PHP > applications and we hope to integrate his code into our RAP - RDF API > for PHP. > > Who of you is attending ESWC? What about a BOF meeting on Tuesday? I'm not although Bastian Quilitz (who has produced a query federator - SPARQL-queries-in, use service descriptions to work out which other services to semnd requests to, rebuld the results). Bastian's at ESWC. Andy > > Chris
Received on Wednesday, 7 June 2006 08:26:49 UTC