Re: Visual Rendering of OWL ontologies?

Hi all,

I've received some replies off-list, particularly some details of the 
UML profiles developed for OWL and RDF as part of the ODM.

One thing I would particularly like to clarify (as Bijan highlighted to 
me) is that this thread is intended to address a visual *notation* for OWL,
not the visualization of ontologies in a broader sense.

I've started a wiki page to capture details of the various Graphical 
Notations for OWL and RDF that have been developed to date:

http://code.google.com/p/owl1-1/wiki/GraphicalNotation

Please contribute any more you are aware of.


I've also recorded some details of the UML Graphical Notation developed 
for RDF and OWL1.0 within the ODM, in a linked page at

http://code.google.com/p/owl1-1/wiki/UMLGraphicalNotation

This info was kindly provided by Elisa Kendall, who was closely involved 
in the development effort.


Best Regards,
Anne


Anne Cregan wrote:
>
> Hi Alexander,
> Please see comments below
>
> Alexander Garcia Castro wrote:
>> My five cents:
>>
>> independent of any ontology editor
>>
>> from my experience the graphical representation of the ontology should
>> not be kept separated from the manipulation of OWL constructs. the
>> graphical representation should also provide the means for direct
>> manipulation. Editors such as Protege have a lot of visual aids, all
>> of them are separated from the editing process. This, in practical
>> terms, means that one always has to go back to the hard-to-manage
>> tree-like-hierarchy. A good example that illustrates the importance of
>> this relationship (editing and visualizing)  comes from very advanced
>> IDEs such as JDEVELOPER, JBUILDER, etc. For these IDEs the
>> visualization facilitates the manipulation of those constructs the
>> language provides, also facilitates the processes one has to run as a
>> programmer.
>>   
> I wasn't meaning in any way to devalue your work - I think it's a 
> fantastic idea to be able to edit
> ontologies graphically, and I'm really looking forward to trying out 
> your stuff!
>
> My motivation was more from the angle of : now we have this ontology 
> and we want to show it to
> people in some generally acceptable way.  I often find myself 
> preparing presentations that show
> ontologies (or at least try to!) and I'd just like to have some 
> consensus on what shape is a class,
> what does a property look like, what does a restriction look like etc, 
> so that people can easily interpret
> what they are seeing in the way they currently can with E-R diagrams 
> for instance.
>
>> do working group members have any strong feeling about a preferred way
>>  
>>> to do it?
>>>     
>>
>> In my opinion this is problem dependent. Not only depends on the
>> ontology at hand, but also the "what do you need the visualization
>> for" affects the choice.
>>   
> That's a good point, and part of the nature and power of ontologies is 
> their ability to be
> viewed and used from many "angles" (by class, by property, by 
> individual etc).  It may well not be
> just one view but several related views that we will need.  It may not 
> be realistic to expect to capture
> a whole ontology in just one diagram.
>
>>  do we want to discuss the approaches and perhaps consider moving
>>  
>>> towards a recommended approach?
>>>     
>>
>> An open discussion may lead to a series of recommendations. that would
>> be nice. I could in that way enrich my plug in, and the rest of the
>> tools we are currently planning to develop.
>>   
> I think something like what Vipul suggested - a UML-like 
> representation - is worth pursuing.
> I look forward to comments from those in the community who have worked 
> on the ODM and
> may have some valuable advice here.
>>  
>>> We already have a task force working an English syntax for OWL1.1,
>>> perhaps we might want to consider
>>> a task force working towards agreeing on a visual representation as 
>>> well.
>>>
>>>     
>>
>> This representation along with the corresponding graphical environment
>> should facilitate the development of ontologies by domain experts.
>> Even if it is just at the level of a baseline ontology.
>>   
> I agree, although IMHO even domain experts should get their hands 
> dirty and use the tools, not
> just look at static diagrams!
>
> Best Regards,
> Anne
>
>>
>>
>> On Nov 10, 2007 4:47 AM, Anne Cregan <Anne.Cregan@nicta.com.au> wrote:
>>  
>>> Hi everyone,
>>>
>>> I have lately seen several approaches to visual rendering of ontologies
>>> (see details below).
>>>
>>> I'm wondering how the group feel about the visual representation of
>>> ontologies as diagrams
>>> independent of any ontology editor:
>>> - do working group members have any strong feeling about a preferred 
>>> way
>>> to do it?
>>> - do we want to discuss the approaches and perhaps consider moving
>>> towards a recommended approach?
>>>
>>> We already have a task force working an English syntax for OWL1.1,
>>> perhaps we might want to consider
>>> a task force working towards agreeing on a visual representation as 
>>> well.
>>>
>>> Thoughts and comments invited.
>>>
>>> Best Regards,
>>> Anne
>>>
>>>
>>> Posting today on CG list from Alexander Garcia Castro 
>>> <alexgarciac@gmail.com
>>>    
>>>> We have developed a new tool that facilitates the generation of
>>>> ontologies in  graphical way. The tool is a plug-in for Protege, it
>>>> uses all of Protege OWL plug in in order to facilitate the direct
>>>> manipulation of OWL constructs. In this way domain experts are able to
>>>> build ontologies in a simple and intuitive manner, the plug-in also
>>>> allows users to load pre-existing ontologies and edit them by using
>>>> the same graphical features. The tool is available at
>>>> http://map2owl.sourceforge.net/, initially our web site is only in
>>>> Spanish, an English version is on the pipe.
>>>>       
>>> There's also a tool called VisioOWL
>>> http://mysite.verizon.net/jflynn12/VisioOWL/VisioOWL.htm
>>>
>>>    
>>>> VisioOWL is a Microsoft Visio application to support the use of Visio
>>>> for creating graphical representations of OWL ontologies. This
>>>> implementation is intended to provide, as close as possible, a direct
>>>> one-to-one mapping between the OWL language constructs and their
>>>> graphical representation. The graphical representation of an OWL
>>>> ontology may provide, for some developers and users, a more
>>>> comprehensive insight into overall class and property relationships
>>>> than could be garnered from the OWL markup alone.
>>>>       
>>> The contact listed is John Flynn jflynn12@verizon.net
>>>
>>> I believe there's also a UML-aligned approach as developed as part 
>>> of the OMG Ontology Metamodel
>>> led by Evan Wallace ewallace@nist.gov
>>>
>>> http://www.omg.org/ontology/
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>     
>>
>>
>>
>>

Received on Sunday, 11 November 2007 07:21:09 UTC