[Date Prev] [Date Next] [Thread Prev] [Thread Next] Indexes: Main | Date | Thread | Author

Re: [ba-ohs-talk] Re: Semantic web meta data

Chris Dent wrote:    (01)

> On Mon, 5 Aug 2002, Murray Altheim wrote:
>>This is the premise of the W3C's Semantic Web effort. I was fortunate
>>enough to sit in on a session at ICCS last year where Terry Winograd
>>(to my mind) demolished the idea of such widespread use of metadata.
> Do you have a reference to a transcript or summary of this?    (02)

Here's a link to the PORT announcement, though this session was not
originally planned so I don't think there are any transcripts. I think
Mary simply managed to pull the four speakers together for an evening.    (03)

   http://www.ksl.stanford.edu/iccs/SemanticWeb.htm    (04)

I do know that the session was videotaped by Mary Keeler, with Jack
Park assisting. Perhaps Jack can fill in some more details.    (05)

----    (06)

The session was a rather impromptu one arranged one of the evenings
by Mary Keeler. It included four speakers, three were faculty at
Stanford. There was Terry Winograd, Deborah McGuiness (who'd recently
made her friendlies with Tim Berners-Lee and the W3C), Geoffrey
Nunberg (in the Linguistics dept. and a regular on NPR's "Fresh Air"),
and one other gentleman (an Irishman, though I'm very sorry to not be
able to locate his name; I have it in my notes at home).    (07)

Deborah McGuiness began with a canned slideshow describing the wonders
of the semantic web. This involved backgrounds showing beaches and ocean
waves. A bit distracting. I can't remember the order exactly, but the
remaining three speakers each in turn took apart the arguments that
might have validated her claims. I believe it was Terry Winograd that
drew a large oval on the blackboard representing the web. This included
all the entertainment, porn, web shopping, personal sites, etc. none of
which would likely ever be organized or contain structured and reliable
metadata. He then drew a very much smaller inner circle being sites that
might use metadata. He divided that three ways and within one of those
horizontal stripes he further divided it, and claimed that only within
that small portion would the vision of the Semantic Web work, only
within those sites that already have structured content, ie., a
database behind them, only sites posted by organizations that would
see value in producing and maintaining such metadata, and only for
those that have customers that *need* such metadata. I'm paraphrasing
horribly (this was last summer), but I think you get the drift: some
very small proportion of the web.    (08)

Geoffrey Nunberg chimed in saying that the state of the art in
computational linguistics was simply not going to assist the process
in any real way, that human language is more complex, more dynamic,
and varies so greatly in context and usage as to make it basically
useless for widescale, "universal" implementation in KM systems. And
this is only for English -- there are hundreds of world languages to
deal with. As a regular radio broadcaster he was very eloquent on the
subject, and I came away feeling that my beliefs going into the
session were correct, ie., that computers are a LONG way from
providing accurate understanding of human speech, a LONG way from
accurately categorizing and organizing anything "globally", and that
general purpose agents are wishful thinking, more likely marketing
propaganda than any reality, designed to maintain research budgets
and fund projects. I suppose this isn't any different than any other
marketing effort, but the Weight of Scientific Inquiry is behind such
ventures in many people's minds.    (09)

As I said, these were my impressions of the talk. I was not impressed
with McGuiness' eludication of the Semantic Web vision, any more than
I was impressed or convinced by Tim Berners-Lee's. The Scientific
American article IMO is a piece of marketing fluff and doesn't have
much of any foundation behind it. A vision, certainly. The Hound of
the Baskervilles was revealed to be a dog covered in phosphorescent
paint. It at least was a real dog, though a mean one. I suppose if
the Semantic Web (being DARPA-funded and apparently backed by
everyone including Bill Gates and Donald Rumsfeld) ever sees the
light of day, it may well bite us in the ass. I'm not in favour of
technologies that further the ability of anyone to keep track of us.
We're already well past 1984 in this regard.    (010)

My apologies for (a) missing that one speaker's name (he was quite
good), and (b) any other errors or omissions.    (011)

Murray    (012)

Murray Altheim                  <http://kmi.open.ac.uk/people/murray/>
Knowledge Media Institute
The Open University, Milton Keynes, Bucks, MK7 6AA, UK    (013)

      If it wants to be a global power and a player in the
      Atlantic alliance, Europe has to get back into the
      business of making war. -- Newsweek Magazine, June 3, 2002    (014)