W3C home > Mailing lists > Public > www-rdf-dspace@w3.org > March 2004

RE: SIMILE PI phone conference, 24-March-04 1100 EDT/1600 BST

From: Butler, Mark <mark-h.butler@hp.com>
Date: Wed, 24 Mar 2004 18:35:15 -0000
Message-ID: <E864E95CB35C1C46B72FEA0626A2E808ED22D3@0-mail-br1.hpl.hp.com>
To: "'David R. Karger'" <karger@theory.lcs.mit.edu>, www-rdf-dspace@w3.org

Hi David

Yes, these are good points. Bijan Parsia, who works with Jim Hendler and was
on the OWL working group visited recently. We had an interesting discussion
about the relationship between deductive databases and persistant
ontologies. Deductive databases were effectively a reformulation of Prolog
which was designed to be more efficient for dealing with large amounts of
persistant data. I asked him for references, and he pointed me at these
papers:

A survey of research on deductive database systems, Ramakrishnan and Ullman
http://www.cs.toronto.edu/~yuana/research/csc2525/93survey-deductive.pdf

Rule Based Languages, Victor Vianu
http://www-cse.ucsd.edu/users/vianu/PAPERS/rules.ps.gz

then a more modern paper relating this stuff to OWL / SW

Implementing OWL Lite in rule-based systems and recursion-enabled relational
DBS, Volz
http://lists.w3.org/Archives/Public/www-webont-wg/2002Oct/att-0033/Paper.pdf

On a more introductory note, I also found Bijan's papers relating RDF to
Prolog at XML.com helpful
http://www.xml.com/pub/a/2001/04/25/prologrdf/index.html
http://www.xml.com/pub/a/2001/07/25/prologrdf.html

cheers, 

Mark Butler
Research Scientist HP Labs Bristol
http://www-uk.hpl.hp.com/people/marbut 


-----Original Message-----
From: www-rdf-dspace-request@w3.org [mailto:www-rdf-dspace-request@w3.org]
On Behalf Of David R. Karger
Sent: 24 March 2004 16:56
To: mark-h.butler@hp.com; www-rdf-dspace@w3.org
Subject: Re: SIMILE PI phone conference, 24-March-04 1100 EDT/1600 BST



Just a stray thought from last week's call. There was much talk about
scalability.  It's important to be clear that there are two completely
different kinds of scalability for us to worry about.  We need database
scalability to deal with large numbers of triples.  Entirely separately, we
need ontology scalability that will let us deal with hundreds of different
ontologies in the system.  The issue of database scalability is relatively
well understood, I think, but the issue of ontology scalability is a
mysterious one.  Our recent demo showed a plausible story for how to cope
with scaling from one ontology to two, but I think we are going to have to
work a lot harder at the UI and system architecture if we want to get all
the way up to hundreds---ie, we can no longer due extensive per-ontology
software development.
Received on Wednesday, 24 March 2004 16:49:35 EST

This archive was generated by hypermail pre-2.1.9 : Wednesday, 24 March 2004 16:49:37 EST