|To:||edbark@xxxxxxxx, Rob Freeman <lists@xxxxxxxxxxxxxxxxxxx>|
|From:||Pat Hayes <phayes@xxxxxxx>|
|Date:||Thu, 31 Jan 2008 20:54:25 -0600|
At 6:20 PM -0500 1/31/08, Ed Barkmeyer wrote:
This is what Rob sent, less the email heads usable only by software.
Perhaps you are filtering this because of another Freeman or "chaoticlanguage.com"
No, it was a much more mundane local problem, now hopefully finally fixed. Sorry.
From: Rob Freeman <lists@xxxxxxxxxxxxxxxxxxx>
Reply-To: [ontolog-forum] <ontolog-forum@xxxxxxxxxxxxxxxx>
General meaningful classes are accessible by clustering words on their
That is a strong claim. Evidence?
Classes found in that way don't have names until you give
Evidence? Can you cite any empirical studies showing that this can be done? (I am aware of the claims of 'latent semantic analysis', which have been applied to large corpora of free texts and have produced some useful results, though nothing like enough to base an ontology on, as far as I know.)
See for instance Hinrich Schuetze's "Dimensions of Meaning":
That paper is a theoretical proposal. Do you have any evidence that this technique has been actually used and has produced useful results? I have seen none.
It may well be that if anyone has achieved significant results using techniques like this they are being kept under proprietary lock and key, as even moderate success could be potentially extremely valuable.
I don't see why it should not prove possible to reason with classes so
Oh certainly, if classes can be identified by any means then it is possible to reason with them, although the results may not be very useful (depending on the classes involved).
I believe indeed that natural language can be thought of as a
The idea that natural language is a formal system is, to put it mildly, not widely accepted among linguists. Certainly, if English is a formal system, it has not so far been formalized.
40 South Alcaniz St.
_________________________________________________________________ Message Archives: http://ontolog.cim3.net/forum/ontolog-forum/ Subscribe/Config: 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 Post: mailto:ontolog-forum@xxxxxxxxxxxxxxxx (01)
|<Prev in Thread]||Current Thread||[Next in Thread>|