Re: [ontolog-forum] Using controlled natural languages for ontology

[Simon Spero <ses@xxxxxxx>]

On Thu, Mar 10, 2011 at 11:58 AM, Ed Barkmeyer <edbark@xxxxxxxx> wrote:

 

<http://attempto.ifi.uzh.ch/site/pubs/papers/doctoral_thesis_kuhn.pdf>

In our experience the problem isn't intelligibility, unless the expressions become extraordinarily convoluted.  The problem is that the
average domain expert naturally _writes_ a different language and takes some training to learn to write the controlled language.  Further, I would add, the domain expert is usually reluctant to 'waste his/her time' doing so.  So the practice is still knowledge engineer working with domain expert to create the ontology.  The primary advantage of using the CNL as a means of _expression_ for _most of_ the ontology is that it allows the domain expert to read, understand and validate that part.  I say 'most of', because there are usually technical considerations in the formulation of the ontology that the domain expert should not be expected to understand -- that is the domain of the knowledge engineer.

 

 [Experts tend to be annoyed when the CNL interpreter complains about what they wrote, especially since its diagnostics only usually identify the syntactic point(s) at which it became confused, and its guidance for what might have been meant is not often helpful.  The worst cases, however, are those in which what the expert writes is unambiguously parsed by the CNL intepreter, but the interpretation it makes is not at all what was intended.  

Ed makes a critical point here. This is the core problem that  Kuhn is addressing in the dissertation as a whole. 

See §2.1.4, which gives an overview of the writability problem -  "One of the biggest problems — if not the biggest problem — of CNLs is the difficulty to write statements that comply with the restrictions of the language.". 

"ACE and its parser were designed according to the error message approach until the development of the ACE Editor (see Section 4.2) began. It simply turned out to be very difficult to provide good error messages. Nevertheless, many CNLs adopt  this error message approach"

Chapter 5, which I mentioned earlier in this thread, is the part of the dissertation which describes the experiments that were done to measure the understandability of the CNL.  Part of the methodology involves representing an ontology graphically. That should be catnip to John Sowa :-)

AceWiki is available at  https://launchpad.net/acewiki , including the predictive editor.  The latest release is version 0.5;  the code is under active development, so the editor has had several improvements since the version described in the dissertation.

Simon

_________________________________________________________________
Message Archives: http://ontolog.cim3.net/forum/ontolog-forum/  
Config Subscr: http://ontolog.cim3.net/mailman/listinfo/ontolog-forum/  
Unsubscribe: mailto:ontolog-forum-leave@xxxxxxxxxxxxxxxx
Shared Files: http://ontolog.cim3.net/file/
Community Wiki: http://ontolog.cim3.net/wiki/ 
To join: http://ontolog.cim3.net/cgi-bin/wiki.pl?WikiHomePage#nid1J
To Post: mailto:ontolog-forum@xxxxxxxxxxxxxxxx    (01)