RE: Tools

>
>
>I've seen a number of largish ontologies generated
>by hand in OWL (and originally DAML+OIL). The problem
>is one of syntax. The XML representation of OWL is
>fairly long winded and hard to read. I'm less than
>convinced that it's appropriate for (editing) a large ontology.
>

Absolutely. Human beings should never have to see XML.

To this end, we have developed an OWL 
viewer/editor called COE which uses an entirely 
graphical GUI based on CMapTools, a system based 
on 'concept maps', ie node-arc-node graphs. The 
CmapTools interface is simple enough for grade 
school kids to use. It takes maybe a day to learn 
the various COE conventions for expressing OWL 
meanings.

COE can only handle medium-sized ontologies (c. 
2K concepts) in its current incarnation - larger 
ones slow down the Java swing graphics rendering 
too much, an issue we plan to work on for the 
next release - and its a little flaky, but feel 
free to play with it. You can get it from 
http://coe.ihmc.us/.

(Go to 'shared ontologies'/IHMC Public Cmap 
Ontologies/Sample Ontologies  to see a whole lot 
of examples that have been rendered using COE, 
all from published OWL ontologies. For a quick 
guide to use, theres a slide presentation at 
http://homam.ihmc.us/modules.php?name=Examples.)

Pat Hayes

>I understand your point about OO style ontologies. I've
>seen this also (probably been guility of it as well).
>But getting people to write long hand ontologies seems
>a less good option than writing tutorials and providing
>good examples of non OO style ontologies!
>
>Perhaps if there were a simpler, more human writable
>syntax, this might be easier, but as it stands writing
>by hand doesn't seem a great thing to me.
>
>Of course, editors have always been a religious
>issue! I'd be surprised if ontology developers shared
>any more consensus on this issue than programmers
>do on IDE's.
>
>Cheers
>
>Phil
>
>
>
>
>________________________________
>
>From: public-semweb-lifesci-request@w3.org on behalf of wangxiao
>Sent: Fri 14/10/2005 18:52
>To: 'public-semweb-lifesci'
>Subject: RE: Tools
>
>
>
>
>-Eric,
>
>>  Helen's point is a very good one.
>>
>>  At the risk of stating what may or may not be obvious to all,
>>  there are several *general* tools that are focused on helping
>>  people create ontologies that may be useful.  In no
>>  particular order ...
>
>Here is my two cents on the topics.
>
>I actually hold a bit different opinion on this.  I think at the beginning
>stage, one should try to do hand editing.  I played around with Protégé
>before, I think because of historic reasons, it uses a lot of terms in
>semantic network.  I saw a lot of people discussing ontologies using "slot",
>"roles", etc. I don't have any grudge on protégé, which I think is a great
>software.  But this sort of dialect is not healthy to advance SW
>technologies. And I also that see many ontologies are developed with an OO
>thinking. Doing it manually actually helps to understand the technology, at
>least that is my experience.  But of course, tool is useful to speed things
>up but only when people knows what the tool are doing for them.
>
>Xiaoshu


-- 
---------------------------------------------------------------------
IHMC		(850)434 8903 or (650)494 3973   home
40 South Alcaniz St.	(850)202 4416   office
Pensacola			(850)202 4440   fax
FL 32502			(850)291 0667    cell
phayesAT-SIGNihmc.us       http://www.ihmc.us/users/phayes

Received on Monday, 17 October 2005 15:45:07 UTC