RE : [OEP,ALL] Potential topics for OEP notes

Fabien Gandon wrotes

> From this weekly practice, I see two notes that could be of interest 
> for these young engineers:
> - a note "from object-oriented design to semantic web modelling" which 
> would summarize and comment shared aspects and core differences; this is 
> probably close to the work on UML.


I'm very interested by a work on this topic. I sent few days ago a few mails  to Alan Rector (Thanks Alan for your support)swhere I Tried to describe the misunderstanding problem for the domain/range subject as it is presented in a the tutorial "Making OWL Easier: Practical Ontology Development in using Protégé-OWL-CO-ODE Tools" where the now famous PIZZA use case example is presented.

Apart the domain/range problem I underlined the fact that, according to our own experience here in FTR&D, there is some cultural blocking points in the OWL comprehension process. Here are some extracts from the mails I sent to Alan :
	"[when FTR&D programmer] see the domain/range problem for pizza_topping and pizza_base, according to their practical knowledge of OOPL strongly typed (JAVA, C++) for example, when somebody define such properties in such a way and when the compiler see the definition of ICE-CREAM-CONE they wait for error-message and in the best case, that's means if they try to understand OWL semantic and the OWA, they really don't understand the [way the reasonner work ]." 
	"After an internal discussion in FTr&D labs, we see the difficulty to understand the Ice_cream_cone example as what we have called here the "OWL paradox". We have understood and are conviced that one of the most important property of OWL is to have a formal semantic which permits, beyond other features, the development of reasoners. So when we read OWL constructs either we have immediatly in mind the exact semantic of the constructs or not. We think that most of such tutorial readers don't have this kw in mind and they only know the OWL standard syntax or the absract syntax or N3. So we are in front of the OWL paradox : we have a language with a clear formal semantic : OWL that for teaching/tutorails needs is presented to the readers with a syntax whom the main property is that, the more the readers are beginners or classical OOL experts and the more the syntax is ambiguous and is interpreted through the personal technical culture filter of the reader (see my very first mail where i say something about OOL Programmers interpretation of OWL constructs and compilation aspects) and so splasch in the hole !!!!!!!"
	I'm convinced that 90% of the FTRD potential users of OWL will use it as an other OOL or a very intelligent version of UML-like languages
Thanks a lot 
Best regards
Marco Nanni


Received on Wednesday, 20 October 2004 08:03:31 UTC