New York Times on the Web Forums Science
Russian military leaders have expressed concern about US plans
for a national missile defense system. Will defense technology be
limited by possibilities for a strategic imbalance? Is this just SDI
all over again?
(7969 previous messages)
rshowalter
- 09:38pm Aug 21, 2001 EST (#7970
of 7982) Robert Showalter
showalte@macc.wisc.edu
Classified information exists within a matrix of correllated
information. That's the statistical foundation on which "latent
semantic analyis" is based. With VERY LARGE masses of correllated
information, connected in VERY MANY WAYS -- there are few "gaps" --
on anything large --- that can really be hidden.
LSA is now fundamental to good search engines - but it is getting
better -- and the correllation procedures people use, in some
significant ways, are better. For that reason, we "make sense of our
world" from INCOMPLETE information.
With computer techniques as they stand, including especially the
web, hiding things is FAR more difficult and risky than it used to
be.
There's a very good paper on LSA -- let me find it. Wish the
search engine was up -- my personal computer search isn't as good.
The logical relation used to break codes -- and do a lot of other
things -- isn't "implication." It is consistency .
. WHAT FITS TOGETHER?
In the world, things have to fit together, and do.
From the open literature, you can zero in on what is possible to
a VERY complete degree.
Because that's true, the Russians ought to be a lot less scared
than they are of missile defense, and we should have sense enough,
right now, with what we know now to know that it is a bad bet
-- and we have to do something else.
Let me do a little searching.
lunarchick
- 09:40pm Aug 21, 2001 EST (#7971
of 7982) lunarchick@www.com
I thought the word missile and defence had been retrospectively
classifed :)
Here's an interesting read brought by a recent visitor to China -
still in wonderment of their game theories.
Author: Ching-Ning
Chu Her game : http://www.fivestarspeakers.com/chu.html
and the book
mine has a silver/black/blue cover.
rshowalter
- 09:44pm Aug 21, 2001 EST (#7972
of 7982) Robert Showalter
showalte@macc.wisc.edu
(While my other computer cranks)
And, on doing LSA-ish stuff, if you work with words, pictures,
and math TOGETHER you can get a lot better consistency
discrimination -- and focus more stuff.
(Also, if Lunarchick is involved, and there is something in the
open literature to be found - the odds are good of finding it -- and
for excellent examples -- look at this thread September 26-28, and
at many other times.)
There is precious little classifiable on missile defense that is
significant to the basic judgements that have to be made. If
anything at all.
lunarchick
- 09:44pm Aug 21, 2001 EST (#7973
of 7982) lunarchick@www.com
shols: www.mamma.com power search input
LSA y white paper
lunarchick
- 09:50pm Aug 21, 2001 EST (#7974
of 7982) lunarchick@www.com
Bing-Fa (soldier - war) The art of war.
221 BC Emperor Chin ordered the burning of all books in China -
and the torturing of scholars. (Is knowledge power?)
2001 Ad Emperor Bush orders the classification of the Webster's
Dictionary and the torturing of selected users.
rshowalter
- 10:01pm Aug 21, 2001 EST (#7975
of 7982) Robert Showalter
showalte@macc.wisc.edu
MD1433 rshowalter
3/24/01 12:38pm
and especially A Solution to Plato's Problem:
The Latent Semantic Analysis Theory of Acquisition, Induction and
Representation of Knowledge by Thomas K. Landauer and Susan
Dumais ..... (Landauer is at the Department of Psychology,
University of Colorado, Boulder, and Dumais is now at Microsoft.)
Here is a draft of that paper, which was accepted with revisions,
and published in Psychological Review , v104, n.2,
211-240, 1997 http://lsi.argreenhouse.com/lsi/papers/PSYCHREV96.html
rshowalter
- 10:15pm Aug 21, 2001 EST (#7976
of 7982) Robert Showalter
showalte@macc.wisc.edu
MD2566-7 rshowalter
4/24/01 8:09pm ... MD2565 rshowalter
4/24/01 7:56pm
MD5353 rshowalter
6/18/01 11:28am . . . MD5354 rshowalter
6/18/01 11:39am MD5355 rshowalter
6/18/01 11:40am ...
I've been wondering for a while. What things are there, now,
that matter with respect to war, peace, or politics, that really
need to be checked, that can't be checked ?
Assuming people are willing to check them.
Not many.
Things are connected together in ways that permits our minds (and
latent semantic analysis) to work to get good patterns very, very,
very often. We form the right patterns from ensembles of
information, considered and organized together. There is plenty of
information around, for checking everything that needs to be
checked. --
The technology for doing it is here, once nation states are
prepared to devote staffs to making sure that they can find out, in
ways that can be crosschecked, the things that matter.
And to insist that the things that can be checked get
checked.
(6
following messages)
New York Times on the Web Forums Science
Missile Defense
|