The subject line of this note is the title of a short article by
Karen Sparck Jones, which I highly recommend: (01)
http://www.sigir.org/forum/2004D/sparck_jones_sigirforum_2004d.pdf (02)
Karen was a pioneer in natural language processing in the late
1950s and remained active in the field until she died in 2007
-- a span of about half a century. (03)
She wrote that article in 2004, and three of the Semantic Webbers,
Nigel Shadbolt, Wendy Hall, and Tim Berners-Lee, tried to respond
to the issues she raised: (04)
http://eprints.ecs.soton.ac.uk/12614/1/Semantic_Web_Revisted.pdf (05)
In my opinion, they did not succeed. (06)
I strongly agree with the conclusion of her article (copy below).
As I have said many times, I believe that logic is essential for
any kind of formal reasoning. But if you are going to do any
kind of informal searching and retrieval that uses something
less rigorous than full logic, there is nothing better than
the ordinary language that people use to ask questions. (07)
The bag o' words used by Google and other search engines is
far better than what we had before, but for complex questions,
it's not as good as asking a question in ordinary language. (08)
John Sowa
__________________________________________________________________ (09)
A discussion contributor reiterated the statement that the SW is
only intended as a bridging device, with no more semantic power
than that needed to provide relatively uncontroversial and hence
common ontological hooks into more specialised domains, and with
ontology concepts defined on pages anyone can read. In attacking
the deep version of the SW I was attacking a strawman. (010)
My argument is not against the practicality and utility of
specialised ontologies, or against the possibility of linking
specialised subdomain ontologies. My argument is that the ontology
approach to the everything that the Web as a whole deals with, once
it goes beyond some very high-level (albeit still semantic) tags
like ‘image’, ‘book’, perhaps ‘news’, which can apply, and be
reasonably similarly understood, across the Web, is a misconceived
enterprise if it assumes that there is something better than natural
language as a general means of expressing, and hence accessing,
information. Indeed the SW notion of ontology definition pages in
natural language is tacitly acknowledging the early limits to formal
rigour. It is much better to bite the natural language bullet right
away. (011)
_________________________________________________________________
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 (012)
|