Re: Direct Mapping

>
> On 7 Sep 2010, at 04:43, Juan Sequeda wrote:
>> On Mon, Sep 6, 2010 at 10:31 PM, Richard Cyganiak
>> <richard@cyganiak.de>wrote:
>>> So there are at least TWO DISTINCT AUDIENCES for the direct mapping
>>> spec:
>>>
>>> 1. RDB2RDF vendors who implement R2RML engines and want to equip
>>> their
>>> systems with functionality similar to D2R's "generate-mapping"
>>> script, which
>>> generates a simple canonical R2RML file for a given database, with
>>> the
>>> intent of allowing further customization of the R2RML file by the
>>> user.
>>>
>>> 2. RDB2RDF vendors who implement RIF-based engines (or engines
>>> based on any
>>> other RDF-to-RDF transformation language). Users of these engines
>>> will write
>>> RIF rules that transform the direct graph into a custom graph.
>>> Users and
>>> vendors of these systems don't need the R2RML language.
>>
>> Great clarification and +1 on everything.
>>
>> Given that we are clear that there are two audience, the only thing
>> I'm
>> saying is that if we are going to write a document on the Direct
>> Mapping,
>> which one of the audiences is expecting to see a R2RML file... we
>> need to
>> have a R2RML syntax before we have a Direct Mapping document ready.
>
> Again, the “RIF-friendly audience” in 2. doesn't want to see R2RML,
> and is better served by a spec that simply describes the shape of the
> resulting direct graph.


Just as a point of realism, there is not a huge or even medium-sized
"RIF-friendly audience" RDF audience :)

 However, Eric's approach does seem interesting and may end up appealing
to lots of people. However, it is clearly focussed on a particular
audience (i.e. RDF people) that is quite smaller than the database
audience. Therefore,  I am happy for there to be a direct mapping
document, although to avoid incoherency the direct mapping document
should ideally be developed so the R2ML spec can hopefully should have a
default option that produces a R2ML document that is coherent with what
the direct mapping document says.

We are chartered to send R2ML to Rec status. If there ends up being a
separate direct mapping document, if people want to keep it separate, can
then either advance to its own Rec status or be produced as a WG note.


>
> The “R2RML-friendly audience” in 1. would be reasonably well served by
> either approach -- describing the shape of the direct graph, or
> describing a canonical “direct R2RML file”.
>
> My preference would be to normatively specify the direct mapping as a
> direct graph. And have an informative appendix that describes an
> algorithm for creating the canonical direct R2RML file that produces
> the direct graph.
>
>> Hence my suggestion to put priority on syntax right now.
>
> I sort of agree. But I don't see a dependency of the direct mapping on
> R2RML.
>
> Richard
>

Received on Tuesday, 7 September 2010 18:00:21 UTC