Since we have been discussing issues of vagueness and precision
in language and ontologies, I decided to mention a recent book by
Kees van Deemter (title above). The author has been working on
problems of semantics in computational linguistics, and the book
is a more popular presentation of issues that he had dealt with
in his more technical papers (cited below). (01)
Following is a book review from the Wall Street Journal: (02)
http://online.wsj.com/article_email/SB10001424052748703808904575025382998649088-lMyQjAxMTAwMDAwMzEwNDMyWj.html (03)
The reviewer quoted some amusing lines that readers of the WSJ
would appreciate: (04)
> "If I seem unduly clear to you," former Federal Reserve Chairman
> Alan Greenspan once remarked, "you must have misunderstood what
> I said." (05)
But the author cites examples from a wide range of language use
and surveys some computational ways of dealing with them. (06)
Following are four technical publications on his web site: (07)
http://www.csd.abdn.ac.uk/~kvdeemte/AC-2009-search.pdf
Vagueness facilitates search (08)
http://www.csd.abdn.ac.uk/~kvdeemte/enlg-invited.pdf
Game theory and language generation: the case of vagueness (09)
http://www.csd.abdn.ac.uk/~kvdeemte/vDeemter-04-final.pdf
Generating referring expressions that involve gradable properties (010)
http://www.csd.abdn.ac.uk/~kvdeemte/vague/vDeemter-inlg04.pdf
Fine tuning natural language generation through experiments with
human subjects: the case of vague descriptions (011)
The basic point of this discussion is that any collection of words from
any source -- Longman's or Wierzbicka's -- will be inherently vague.
They might be useful as rough guidelines (note the paper about the
way vagueness can facilitate search), but they won't be sufficiently
precise for any kind of formal specification. They could never ensure
that two or more programmers would interpret them in the same way. (012)
John (013)
_________________________________________________________________
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 (014)
|