W3C home > Mailing lists > Public > semantic-web@w3.org > December 2011

Re: Turtle/UML

From: Sergio Fernández <sergio.fernandez@fundacionctic.org>
Date: Wed, 21 Dec 2011 10:28:19 +0100
Message-ID: <CAMR8oe1pjbYGXtBvS5iUwYmbFhjvheFYM=-4ZBeakjV_A0yOCQ@mail.gmail.com>
To: Cory Casanave <cory-c@modeldriven.com>
Cc: "Andrea Splendiani (RRes-Roth)" <andrea.splendiani@rothamsted.ac.uk>, Deepak Nadig <deepak.nadig@gmail.com>, "<semantic-web@w3.org>" <semantic-web@w3.org>
Taking into account the differences between both models (RDF vs. OOP),
the representation of properties doesn't fit well as attributes on a
UML diagram. That's why some years ago on Neologism
(http://neologism.deri.ie/) we preferred to generated an ad-hoc
diagram. Maybe it's time to launch an initiative to get an standard
graphical representation of the RDF model...

On 21 December 2011 00:32, Cory Casanave <cory-c@modeldriven.com> wrote:
> We do the opposite – we generate RDF and OWL from UML models.  We find the
> UML tools much more mature (Our favorite is Magicdraw) and the notation more
> stakeholder friendly.  We also generate other artifacts (code, descriptors,
> XML, etc) from the same models which include other system aspects such as
> services and processes. We could easily gen turtle.  We use UML extended a
> bit with profiles but it is mostly normal UML, for full-boat OWL more
> extension to UML is needed but most of our requirements don’t call for
> full-boat OWL.  The UML subset used for LoD seems at about the right
> semantic level for many applications.
>
>
>
> Regards,
>
> Cory Casanave
>
>
>
> From: Andrea Splendiani (RRes-Roth)
> [mailto:andrea.splendiani@rothamsted.ac.uk]
> Sent: Sunday, December 18, 2011 2:47 PM
> To: Deepak Nadig
>
>
> Cc: <semantic-web@w3.org>
> Subject: Re: Turtle/UML
>
>
>
> Hi,
>
>
>
> Turtle is a "low level" language. From a generic you are unlikely to derive
> meaningful UML.
>
> You could imagine than a fragment of Turtle will contain a few rdf:type or
> x:subClassOf expressions and hence try to model these as UML
> classes/subclasses. And you could use property assertions to derive
> properties in UML.
>
> In this case what you would be doing would be to derive UML from rdf/owl/x
> properties (which can be expressed in turtle, as in other syntaxes).
>
> While this seems a straightforward thing to do, it may turn out to be
> painful if you are not clear about your application case. There are no
> schemas or models in RDF/Turtle, and when you see classes and properties,
> they are not the same thing you think of in UML.
>
> Modeling in UML and serializing in RDF/Turtle is more doable.
>
>
>
> my 2 cents,
>
>
>
> best,
>
> Andrea
>
>
>
> Il giorno 16/dic/2011, alle ore 17.26, Deepak Nadig ha scritto:
>
>
>
>
>
> Are there any tools that can generate UML class/object diagrams from Turtle?
>
>
>
> Thanks,
>
> Deepak
>
>



-- 
Sergio Fernández
CTIC - Technological Center
Parque Científico y Tecnológico de Gijón
C/ Ada Byron, 39 Edificio Centros Tecnológicos
33203 Gijón - Asturias - Spain
Tel.: +34 984 29 12 12
Fax: +34 984 39 06 12
E-mail: sergio.fernandez@fundacionctic.org
http://www.fundacionctic.org
Privacy Policy: http://www.fundacionctic.org/privacidad
Received on Wednesday, 21 December 2011 09:28:55 UTC

This archive was generated by hypermail 2.4.0 : Thursday, 24 March 2022 20:41:30 UTC