Re: RDFSync 0.1

Hi Giovanni,
This seems like a very interesting technological development.
It has wide applicability.
I would like to ask a couple of questions.
1. Demonstrators of ubiquitous computing such as COBRA from UMBC
http://ebiquity.umbc.edu/project/html/id/1/
use their architecture which includes agents that encapsulate OWL rule sets
to mediate those rules. How does this tecfhnology touch on that sort of
usage?
2. Quite separately I am also interested in re-postioning XML -- > RDF
(OWL), actually more acurately XSD -> RDFS/OWLS. There are some tools and a
general theory (OWLmt, hmafra, Model Management - graph theory).
How does this touch on that topic, because it seems to me that it does.
Isn't XMl -> RDF (more general, less well defined) -> (more specific, less
ambiguous) where there is possible information loss in the resolution of the
ambiguity and where different RHS (RDF)  expressions may be derived from the
LHS (XML)?
Wouldn't versioning and version tracking have a role to play in this since
alternatives are different versions that maybe proposed as the "same" but
without loss of underlying data, as I understand what you are working on. I
am sure this can be pushed further in terms of automating the process of XML
-> RDF?
Any thoughts?
Best,
Adam
 <http://ebiquity.umbc.edu/project/html/id/1/>

On 22/03/06, Giovanni Tummarello < g.tummarello@gmail.com> wrote:
>
>
> The problem is really general.. but there have been many mentioning
> applications for one such procedure. Not last look at the early postings
> by TBL and others referenced in the paper.
>
> Anyway, scenario include but are not limited to scalable cluster of RDF
> DBs (powering similarly scalable web sites, semantic web services)
> especially in loose heterogeneous, P2P or grid environments; mobile
> devices syncing.
> One for all, we will use it right away in DBin version 0.4 coming soon (
> http://dbin.org ): RDFSync is used each time one client meets another in
> P2P group earlier, the entire graph surrounding a URI had to be passed
> at each p2p direct exchange. Not good as the data increased in the group.
> Giovanni.
>
> Hans Teijgeler wrote:
> > Hi Giovanni,
> >
> > Interesting, but what requirement is fulfilled by this?
> >
> > Regards,
> > Hans
> >
> > -----Original Message-----
> > From: semantic-web-request@w3.org [mailto:semantic-web-request@w3.org]
> On
> > Behalf Of Giovanni Tummarello
> > Sent: Tuesday, March 21, 2006 23:53
> > To: semantic-web@w3.org
> > Subject: RDFSync 0.1
> >
> >
> > RDFSync 0.1
> > -----
> >
> > Introduction
> >
> > Remote RDF model synchronization is a procedure by which one efficiently
> > updates a local model based on a remote one. By updating one can mean
> "make
> > identical to the remote" (classic RSync) but in this case it might mean
> > "patch" with information from a remote model not present in the current
> one.
> >
> > Classic procedures such as RSync cannot be applied to RDF models since
> they
> > usually serialize nondeterministically. Even if a deterministic
> > serialization algorithm such as is used (Such as J. Carrol's canonical
> > serialization of RDF Graphs) , the performance of classic RSync remain
> poor,
> > as minimal modifications in the model might cause completely different
> > canonical serializations.
> >
> > RDFSync is an efficient RDF Synchronization algorithm based on the
> canonical
> > serialization of a Minimum Self Contained graph decomposition of remote
> RDF
> > models. For further details, experimental results and demo code please
> > visit:
> >
> > http://semedia.deit.univpm.it/tiki-index.php?page=RdfSync
> >
> > as highlighted in the short paper, some issues do remain. Looking
> forward
> > for direct feedbacks from those interested.
> >
> > Giovanni Tummarello
> > also on behalf of Christian Morbidoni and others at Semedia -
> > http://semedia.deit.univpm.it
> >
> > --
> > No virus found in this incoming message.
> > Checked by AVG Free Edition.
> > Version: 7.1.385 / Virus Database: 268.2.6/286 - Release Date: 20-Mar-06
> >
> >
> >
>
>
>