W3C home > Mailing lists > Public > www-rdf-interest@w3.org > September 2002

OpenSource release of Parka, University of Maryland Database backend (fwd)

From: Ron Alford <ronwalf@wam.umd.edu>
Date: Fri, 20 Sep 2002 15:16:54 -0400 (EDT)
To: www-rdf-interest@w3.org
Message-ID: <Pine.LNX.4.44.0209201515570.15192-100000@cerebrus.volus.net>



I'm forwarding this along from Dr. Hendler, since his tries weren't going 
through.
--
Ron Alford, KB0NUV
http://volus.net/~ronwalf


---------- Forwarded message ----------
Date: Fri, 20 Sep 2002 12:54:43 -0400
From: Jim Hendler <hendler@cs.umd.edu>
To: rdf-interest@w3.org
Cc: semweb@mindswap.org
Subject: OpenSource release of Parka, University of Maryland Database backend

(Apologies if you get this twice - URIs were stripped off previous sending)

We are pleased to announce the opensource release  [1] of the 
(no-longer patent-pending) Parka system.  The new version, ParkaSW, 
is aimed at supporting RDF use by being a scalable backend technology 
for the storage and retrieval of RDF graphs.
  Parka's representations are still more tailored for the earlier 
SemWeb language SHOE [2], but it also makes a capable RDF tool - see 
[3] for example.
Parka's inputs are triples, and the builder file can easily be 
created from RDF.  Parka also includes inheritance algorithms and 
others that will be of use for OWL-lite and RDFS users, however they 
currently require some preprocessing algorithms.  We hope to release 
a new version in the next few months which will more directly support 
RDFS and OWL-lite.
  The current implementation makes ParkaSW unable to contain more than 
about 2.5E6 (2.5 million) triples, we are also working on a new 
version that will remove that limitation.
  Parka currentl benchmarks faster than other RDF databases due to its 
innovative way of storing triple tables.  Note, however, that for RDF 
DBs that have many different properties (more than 300) Parka can run 
up against file limitations on some Unix implementations - thus to 
handle a large number of relations in a single graph, some parameters 
needs to be munged.  Contact us if you get into that situation (most 
DBs we've tested on have far less properties than that in any graph).
   -Jim Hendler
   for the MINDSWAP research group (http://www.mindswap.org)

note:  ParkaSW derives in part from Parka-DB(TM) which is still 
available for licensing - see the details on the release page [1].

[1] http://www.mindswap.org/2002/parka
[2] http://www.cs.umd.edu/projects/plus/SHOE
[3] http://www.cs.umd.edu/projects/plus/Parka


-- 
Professor James Hendler				  hendler@cs.umd.edu
Director, Semantic Web and Agent Technologies	  301-405-2696
Maryland Information and Network Dynamics Lab.	  301-405-6707 (Fax)
Univ of Maryland, College Park, MD 20742	  240-731-3822 (Cell)
http://www.cs.umd.edu/users/hendler
Received on Friday, 20 September 2002 15:37:52 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Monday, 7 December 2009 10:51:56 GMT