Re: Leveraging UML in RDF

Hi Sergey, 

you might be interested in the XML Metadata Interchange (XMI)
to adopt the names for your RDF-classes and -properties describing
an UML model. 

- http://www.oasis-open.org/cover/xmi.html
- ftp://ftp.omg.org/pub/docs/ad/98-10-16.txt

Reinhold

<cit>
XMI 1.1 RTF Main revised document - OMG XMI 1.1 RTF. 25 October 1999.
OMG XMI v. 1.1: Revisions and Recommendations Summary. By: OMG XMI
Revision Task Force. OMG Document ad/99-10-02, 284 pages. "The main   
purpose of XMI is to enable easy interchange of metadata between
modeling tools(based on the OMG UML) and metadata repositories (OMG MOF
based) in distributed heterogeneous environments. XMI integrates three
key industry standards: (1) XML - eXtensible Markup Language, a W3C
standard; (2) UML - Unified Modeling Language, an OMG modeling standard;
(3) MOF - Meta Object Facility, an OMG metamodeling and metadata
repository standard. The integration of these three standards into XMI
marries the best of OMG and W3C metadata and modeling technologies,
allowing developers of distributed systems to share object models and
other metadata over the Internet." 
</cit>

Sergey Melnik wrote:
> 
> Folks,
> 
> the UML community developed a set of useful models for representing
> static and dynamic components of software-intensive systems. UML is an
> industry standard and serves as a modeling basis for emerging standards
> in other areas like OIM, CWM etc. As of today there exist a variety of
> UML vocabularies for describing object models, datatypes, database
> schemas, transformations etc.
> 
> Wouldn't it be nice to make these models universally accessible,
> evolvable and mixable on the Web? This is not a purely altruistic
> statement. In fact, I needed a comprehensive finite-state machine model,
> and UML has got a really nice one that I want to reuse.
> 
> I'm working on making UML "RDF-compatible". The first step is to
> represent the UML conceptual model in RDF. This is similar to defining
> an alternative RDF Schema specification. UML bootstraps itself in a way
> similar to RDFS. It is however significantly more verbose and contains a
> lot of details that go far beyond RDFS.
> 
> I gave it a try. RDF-encoding of the UML Foundation/Core package
> (without "auxiliary elements") can be found at
> 
>         http://www-db.stanford.edu/~melnik/rdf/uml/uml-core.rdf
> 
> I don't have a Web page describing the rationale, and pros and cons of
> the design yet. Have a look. If you are familiar with RDFS, you'll
> probably understand the specs without knowing much about UML. [1] and
> [2] can be of help. I'm using only the basic RDF constructs such as
> rdf:type and rdf:Seq. The encoding is incomplete and may contain
> "semantic" mistakes.
> 
> Higher-level UML models (like statecharts) can be encoded using the
> UML/RDF vocabulary defined on the above page. Let me know what you think
> of the idea.
> 
> Sergey
> 
> [1] http://www.w3.org/TR/NOTE-rdf-uml/
> [2] ftp://ftp.omg.org/pub/docs/ad/99-06-08.pdf

Received on Saturday, 22 April 2000 09:13:52 UTC