Re: Role of the Ontology and Expressivity - to discuss on telcon

On Tue, Apr 27, 2010 at 7:36 AM, Eric Prud'hommeaux <eric@w3.org> wrote:

> * Juan Sequeda <juanfederico@gmail.com> [2010-04-26 20:27-0400]
> > This week I commented on the role of the ontology [1] where we should
> > consider that we are mapping to the following:
> >
> >    - existing domain ontology (FOAF, SIOC, etc)
> >    - a putative ontology (automatically generating the ontology from the
> >    schema/DDL)
> >    - a federated ontology: combining different ontologies
> >
> >
> > Eric initially manifested the need to talk about the expressivity [2]. If
> I
> > understand correctly the issue of the expressivity is being presented as
> > "DIRECT - Recapulating Relational Structure" [3] and TRANSFORM -
> > Non-isomorphic transformation [4].
> >
> > I'm honestly a bit confused with both of these requirements, but this is
> my
> > take on them. Eric, please correct me if I'm wrong:
> >
> > DIRECT: What I understand is that you could essentially automatically
> > generate RDF triples (a graph) from the relational data. The top image
> shows
> > a graph of relational data and the bottom image shows how the previous
> graph
> > can be exposed as RDF (a graph) and it is equivalent (isomorphic) to the
> > relational data graph.
> >
> > TRANSFORM: In this section you have taken the previous relational data
> graph
> > and put it into the graph structure of two different ontologies. You show
>                                                       ^^^^^^^^^^
> > that if you compare the relational data graph with the two graph
> structures
> > of the two different ontologies, then the graphs are different
>                       ^^^^^^^^^^
> > (non-isomorphic).
>
> I believe that you understand the expressivity points DIRECT and
> TRANSFORM, but I don't believe we can use the word "ontology" to
> describe graph shape. For example, an attribute diabetes_history
> could be represented in the SNOMED ontology as
>  # likely isomorphic to relational graph.
>  _:p sn:160303001 true .       # family history of diabetes
> or as
>  # likely non-isomorphic, but better model.
>  _:p sn:57177007 _:h .         # family history
>  _:h sn:24609004 sn:73211009 . # with finding of diabetes.
> As a sense of scale, the most popular medical ontology, SNOMED,
> has around 370k analogous representation choices (pre-coordinated
> terms).
>
> Pre-existing ontologies also demand term synthesis. A ball bearing
> database with hardening treatment T5 would likely be expressed in a
> pre-existing RDF graph as a URI in an ontology of eutectic hardening.
>
> In this, pre-existing ontologies are a motivation for a set of
> expressivities.
>

This is a huge issue and I'm totally on the same page as you.

In our case, we work with biologist and most of their ontologies are in OBO
which can be translated to OWL. It is very hard to map the relational schema
to these ontologies, but we have to because these are the ontologies that
have been developed explicitly for interoperability.

I still do not understand why you suggest that we don't use the word
ontology (or even vocabulary). Isn't that what we are actually doing,
mapping to an ontology?


>
> > IF I UNDERSTOOD THIS CORRECTLY... then
> >
> > What I was mentioning initial (putative ontology and domain ontology) is
> > directly aligned to what Eric is presenting. Essentially, creating a
> > putative ontology directly from the sql ddl is equivalent to the DIRECT
> > requirement and mapping the rdb schema to an existing domain ontology is
> the
> > TRANSFORM requirement. Therefore, I propose that this needs to be
> reworded.
> > Honestly, it is very difficult to understand and I think the semantic web
> > and database community would both understand that we are talking about
> > mapping to an ontology.
>
> I agree that this is a difficult point, but I don't think that we can
> gloss over it by describing our requirements in terms of motivations
> as it doesn't give the required precision. I'd of course, love
> clarifications and refinements though, as it is difficult to convey
> graph theory in a paragraph.
>
>
> > Sometimes there doesn't exist a domain ontology that I would like to map
> to,
> > therefore I will just use the DIRECT approach. This is what I consider
> > "direct mapping", therefore we are totally aligned in our thoughts. On
> the
> > other hand, you do want to map to existing domain ontologies. These
> mappings
> > will have a higher level of complexity.
> >
> > In conclusion, we I propose that we present these requirements as:
> >
> > The mapping language should be able to expose the relational data as RDF
> > without considering a domain ontology
> >
> > The mapping language should be able to express the mapping between the
> > relational schema and a existing domain ontology
> >
> > [1]
> http://lists.w3.org/Archives/Public/public-rdb2rdf-wg/2010Apr/0059.html
> > [2]
> http://www.w3.org/2001/sw/rdb2rdf/wiki/Draft_of_Use_Cases#Expressivity
> > [3] http://www.w3.org/2001/sw/rdb2rdf/use-cases/#DIRECT
> > [4] http://www.w3.org/2001/sw/rdb2rdf/use-cases/#TRANSFORM
>
>
> I intended to address this point in
> <
> http://www.w3.org/mid/m2z2eb0b7511004230440qfe6655d9x28f4c1e220972476@mail.gmail.com
> >
> summaried below.
>
>
I still find this hard to understand, and I'm somebody who understands the
issue. So I'm guessing that this may be difficult for non tech people to
understand. I suggest that we have to rewrite this.s


>
> Date: Fri, 23 Apr 2010 07:40:34 -0400
> From: Eric Prud'hommeaux <eric@w3.org>
> To: Juan Sequeda <juanfederico@gmail.com>
> Subject: Re: The Role of the Ontology
> Message-ID: <
> m2z2eb0b7511004230440qfe6655d9x28f4c1e220972476@mail.gmail.com>
> * Eric Prud'hommeaux <eric@w3.org> [2010-04-23 07:40-0400]
> > I've added
> > [[
> > This graph can be used when it is desired to let the database
> > structure determine the effective ontology of the RDF view.
> > ]]
> > to <http://www.w3.org/2001/sw/rdb2rdf/use-cases/#isomorphic> and
> > [[
> > It is good Semantic Web practice to use shared ontologies. Using
> > popular ontologies, or even ontologies which represent information in
> > multiple domain databases, usually requires transformations
> > [GraphTransform] to the direct graph.
> > ]]
> > to <http://www.w3.org/2001/sw/rdb2rdf/use-cases/#non-isomorphic>.
> >
> > I've also added an extensive editorial note outlining the requirement
> > space implied by the later "requirement":
> > [[
> > There are many types of graph transformation, representable by SQL
> > views, SPARQL CONSTRUCTs, Horn logic, etc. The RDB2RDF workiing group
> > would like feedback to decide if, for instance, the transformation
> > expressed by RIF Basic Logic Dialect are appropriate. Use case
> > contributions to this document will help determine the exact
> > expressivity required.
> > ]]
>
> --
> -ericP
>