W3C home > Mailing lists > Public > public-owl-dev@w3.org > April to June 2009

Re: UML-OWL Generator, A product to convert UML into OWL

From: Bijan Parsia <bparsia@cs.manchester.ac.uk>
Date: Sun, 3 May 2009 17:26:16 +0100
Cc: <public-owl-dev@w3.org>, "'Elisa F. Kendall'" <ekendall@sandsoft.com>
Message-Id: <C0DDFB87-0461-48CB-8892-73FEAD687CEB@cs.manchester.ac.uk>
To: "UML-OWL Gen" <info@umlowlgen.com>
On 3 May 2009, at 17:01, UML-OWL Gen wrote:

> Bijan,
>
> We'll take your sarcastic feedback in positive way, even it wasn't.

That's big of you! I'm really really really impressed!

(Btw, *that* was sarcasm. There was no sarcasm in my original post.  
Accusing someone of sarcasm is a good way to end up with a big heaping  
dose of it :))

> The generator is "Patent Pending", so you will not find in Google  
> for a year
> period.

Ah. That makes your link to usa.gov confusing.

A white paper or some other technical description would be helpful.

> This work is different from Elisa's work, where her work is more  
> about ODM,
> UML-OWL Generator has nothing to do with ODM.

I don't know if having something to do with ODM matters.

> The generator works only from UML to OWL, and the unique thing about  
> it is
> how it deals with complex UML models & metamodels of multiple  
> packages,

How exciting! You must be very excited! It's so exciting!

> not
> toy examples as most literatures ended up with.

Again, it's hard to evaluate anything when the concrete technical  
details are so lacking. It's hard to evaluate the spread of your  
patent as well when there are no details.

But I'm confused by what the problem *would* be. Either there is a  
mapping from UML into OWL or there isn't. (There are many of course,  
but I generally think of the ones that purport to capture the  
semantics of UML diagrams). Once you have the mapping, implementing  
such a transform is, as long as it's purely syntactic, trivial. (In  
the sense of a mere matter of parsing.)

So the only part that seems, afaict from the extremely limited  
information supplied, to be *possibly* novel is the mapping itself.  
And yeah, I'm skeptical that that is interestingly novel.

> We would be happy if you point us at any product of one of "many
> literatures" that works in real production environment.

This seems to confuse implementation with invention. It also confuses  
the burden of proof. If *you're* seeking a patent, as  
you ...er...*are*, then presumably *you've* done the relevant  
literature search and analysis. Thus, you should be in a position to  
1) tell me what the alternatives are and 2) tell me the differences  
between them and your approach. If you have, in fact, missed something  
that I know about, I'll say so. Similarly, if you've misunderstood  
something, and I detect it, then I'll say so as well.

It doesn't look good when you don't provide technical details. Maybe  
you have to not, but that doesn't make it good looking :)

> We appreciate your feedback on the website,

Always glad to help.

> we will add a web service for
> "Try It" to support interactive test, and will make models and  
> ontologies
> available on "Test Case".

Those will make it easier to figure out what you're doing. A brief  
white paper would also be helpful. I encourage you to submit a  
description to e.g., OWLED.

Also, I'd be happy to know your name :)

I'd *love* to have a good story for OWL and UML, but thus far, you're  
not selling it to me :)

Cheers,
Bijan.
Received on Sunday, 3 May 2009 16:26:59 GMT

This archive was generated by hypermail 2.3.1 : Wednesday, 27 March 2013 09:32:57 GMT