Re: Reverse Mapping RDF2RDF

On Sun, Nov 21, 2010 at 9:31 AM, Alexandre Bertails <bertails@w3.org> wrote:

> [ snip ]
> > > The relation between RDB, RDF, SPARQL and SQL should be formalized and
> > > should not rely only on examples. It's called semantics preservation
> > > -- a well-known problem in the compilation field -- and IMO it should
> > > be in a normative section.
> > >
> > > Eirc and I are totally confident that in the case of the direct
> > > mapping, we can map any SPARQL query to its equivalent SQL query.
> >
> > Where is the proof?
>
> heh :-)
>
> I think we first need to have a definitive version of RDB + Direct Mapping.
>

I don't think the proof doesn't depend on the "definitive version".
Essentially there is a concept of direct mapping a RDB to RDF. The proof
must work, regardless the way you represent it (sets, rules).

The proof is actually simple if you consider the direct mapping rules.
Angles and Gutierrez proved that "SPARQL and non-recursive Datalog with
negation have the equivalent recursive power, and hence, by classical
results, SPARQL is equivalent from an expressive point of view to Relational
Algebra". By representing the mapping in datalog rules, one can represent
SPARQL as a datalog query, and then use the mapping rules to rewrite the
query to relational algebra.

Therefore, thanks to the Angles and Gutierrez result, we don't have to worry
about the SPARQL to SQL issue, from the direct mapping perspective. For
R2RML, if that language has more expressivity than datalog, then we might be
in trouble.



> The first step is to be sure that this process is complete. For RDB,
> this is done by definition, through its abstract model. The proof that
> the direct mapping can react on *any* RDB instance is done
> automagically -- well, not so magic :-) -- by Scala and its
> type-checking.

We are still working on sparql2sql but you can already play with what
> we have [1] (usable but not complete). The "confidence" I was talking
> about comes from the fact I've done that kind of job (formal proof
> with compilers) in the past. Actually, the chosen formalism reflects
> this concern.
>
> But again, this is out-of-scope. I nevertheless support Ivan's idea of
> a WG note on this subject, but only *after* the direct mapping will be
> out there, and *complete*.
>
> Alexandre.
>
> [1] http://dvcs.w3.org/hg/FeDeRate/
>
> >
> > >
> > >> I scare-quote equivalence because SPARQL returns RDF terms and SQL,
> unless you significantly change the wire protocol, returns strings which can
> be parsed to RDF terms.
> > >
> > > I don't understand, a SQL result is a SQL table, not a bunch of
> > > strings to be parsed. It's perfectly defined (the same way than for
> > > the relation algebra).
> > >
> > > Alexandre.
> > >
> > >>
> > >>> That is all...
> > >>>
> > >>> Ivan
> > >>>
> > >>> On Nov 19, 2010, at 20:20 , Juan Sequeda wrote:
> > >>>
> > >>>> Hi Ivan,
> > >>>>
> > >>>> Per the agenda, it states:
> > >>>>
> > >>>> 4. Reverse Mapping
> > >>>> Question from Ivan re RDF2RDB (it's not in our charter, but maybe
> some WG
> > >>>> members plan to address this?)
> > >>>>
> > >>>> I'm curious about RDF2RDB. Could you expand on this. What are the
> use cases? Requirements?
> > >>>>
> > >>>> Thanks
> > >>>>
> > >>>> Juan Sequeda
> > >>>> +1-575-SEQ-UEDA
> > >>>> www.juansequeda.com
> > >>>
> > >>>
> > >>> ----
> > >>> Ivan Herman, W3C Semantic Web Activity Lead
> > >>> Home: http://www.w3.org/People/Ivan/
> > >>> mobile: +31-641044153
> > >>> PGP Key: http://www.ivan-herman.net/pgpkey.html
> > >>> FOAF: http://www.ivan-herman.net/foaf.rdf
> > >>>
> > >>>
> > >>>
> > >>>
> > >>>
> > >>
> > >>
> > >>
> > >
> > >
> > >
> > >
> >
>
>
>
>
>