|
On 5/22/2013 5:03 PM, David Eddy wrote:
Ed -
On May 22, 2013, at 4:54 PM, Ed Barkmeyer wrote:
Creating a
well-defined CNL with a well-defined logical
interpretation,
Where does one find this beast, CNL?
It's been my experience that most organizations—notice no
definition of size here—don't have a clue what their language
is, much less control it.
I'm still fighting the insanity of "We're going to have a
single name for a thingy across the enterprise..."
David,
A 'controlled natural language' is a grammar. Some CNLs also
legislate the vocabulary; most simply provide a means for the user
to supply the vocabulary. Many CNL vocabularies support synonyms --
the thingy can have multiple names. Whether the association between
a vocabulary term and its meaning is fixed or contextual is
(usually) a feature of the CNL. Some simply force the context to be
explicit by having 'compound noun' (a context term followed by the
noun it restricts) as a grammatical construct in the language.
You should look at Tobias Kuhn's paper. Its scope is impressive.
http://attempto.ifi.uzh.ch/site/pubs/papers/kuhn2013cl.pdf
The purpose of a CNL is not to be "the language of the enterprise".
It is rather a formal language to be used for capturing and
conveying specific kinds of information for some technical usage.
The objective of a CNL is to be clear and unambiguous, while being
readable by practitioners in some domain.
A good example (which Tobias discusses) is Simplified Technical
English, which was originally developed to allow aircraft engineers
to communicate clear requirements to the guys who build the parts.
STE does not have a grammar per se: it makes strong suggestions
about what English grammatical structures should be used when, and
what ones should not be used. It provides a base vocabulary, with a
rule: Use the word X when you mean X, but it allows the user
discipline to add its own specific terms. Many STE requirements
sentences have a clear logical formulation, assuming appropriate
relations. Others use English conventions that are clear to human
readers, but require a lot of careful thinking for a first-order
logic rendition. No one intends STE to be the lingua franca of any
enterprise. It is recommended practice for writing engineering
requirements, full stop. And that kind of constrained usage is the
purpose of most CNLs. A further goal of the more recent CNLs is to
have a clear rendering into some formal logic, or at least some
formal modeling language.
-Ed
__________________
David Eddy
--
Edward J. Barkmeyer Email: edbark@xxxxxxxx
National Institute of Standards & Technology
Systems Integration Division, Engineering Laboratory
100 Bureau Drive, Stop 8263 Tel: +1 301-975-3528
Gaithersburg, MD 20899-8263 Cel: +1 240-672-5800
"The opinions expressed above do not reflect consensus of NIST,
and have not been reviewed by any Government authority."
|
_________________________________________________________________
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 (01)
|