Ferenc, (01)
I agree with that point: (02)
FK> The current theoretical approach to semantic analysis as used
> in AI or applied MT is certainly nothing to write home about,
> especially, if you look at the actual products in contrast to
> ambitions. (03)
But what I was objecting to is the fantasy land of testing,
which pretends that just because there is a number attached to
a test score that the number represents some kind of reality. (04)
FK> In the practical world of business, such as in car manufacturing,
> there are proven methods of giving account of multitudes as well
> as unary objects. Moreover, they have a way to assemble them
> with methods to measure the number of mistakes/faults/rejects
> made in the process which are crucial as feedback to improve
> the methods applied. (05)
I also agree with that point. (06)
The invention of money was a revolutionary advance that made it
possible to quantify business transactions. The bankers in the
Italian Renaissance invented very precise methods for doing
bookkeeping. My former employer, the International Business
Machines Corporation, became very prosperous by making punched
card equipment for keeping track of the numbers, and they
became even more prosperous when they replaced the punched
cards with electronic counters. (07)
But the fact that it's possible to keep track of the money
with great precision does not mean that there is a scientific
method for putting an exact value on the goods that are sold. (08)
The traditional ratio of the price of gold to the price of silver
has been 12 to 1. That is why there are twelve troy ounces in
one troy pound: one ounce of gold is worth one pound of silver.
And the reason for that ratio is that gold represents the sun,
silver represents the moon, and there are 12 months in 1 year. (09)
We are still in the year-end holiday sales period, which is
critical to the profitability of most retail businesses.
But if you look at the kinds of prices and discounts, you
see that each store is trying to out-guess the customers and
the other stores about what kind of gimmick might bring in
the most cash. In fact, for luxury items, they often discover
that *raising* the price increases sales. The sun and moon
begin to look like a more rational basis for determining price. (010)
In short, I agree with you that measuring is good when you can
measure something real. I also agree that a lot of semantic
analysis in AI and software engineering is more fantasy than
reality. And I am warning people that trying to measure that
fantasy can be very bad -- because some people will think that
the numbers they get represent reality. (011)
John (012)
_________________________________________________________________
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 (013)
|