DM: Datamining Definition

Datamine-L didn't post this. Maybe it is more relevant to
Machine-Learning. The definitions of datamining and machine-learning
converge it seems. Any differences that anyone can see?
FWP.

---------- Forwarded message ----------
Date: Mon, 10 Apr 2000 11:17:22 -0700 (PDT)
From: Franklin Wayne Poley <culturex@vcn.bc.ca>
Reply-To: Robot-for-President@egroups.com
To: datamine-l@nautilus-sys.com
Cc: robot-for-president@onelist.com
Subject: [Robot-for-President] re DM: Datamining Definition.

-----Original Message-----
From: Frank Buckler <buckler@m2.uni-hannover.de>
To: datamine-l <datamine-l@nautilus-sys.com>
Date: Tuesday, April 04, 2000 1:24 PM
Subject: AW: DM: Datamining Definition.

clip

 >So here is my suggestion:
 >
 >DM is: "Acquiring knowledge (laws and regularities) automatically from data
 >in order to achieve an business objective of an enterprise."

I like this as a definition for machine learning in general. As this very
interesting discussion expanded I had to wonder if DM wasn't leading to
that. Kurzweil doesn't use the expression "data mining" in his latest book
(1999) but he does describe some specific aspects of intelligent machinery
which would be right on topic for this list. For example he talks about
NN's and evolutionary algorithms in the context of software which
substitutes for human financial analysts, noting that some 5% of stock
investments are now made this way (p. 87). Closer to the issue of
definitions he says, "In the next decade, we will see...computerized
personal assistants that can converse and rapidly search and understand
the world's data bases...." (p.4). Now that is a very optimistic and
somewhat surprising statement. But given that this is one of the most
eminent contributors to the development of intelligent machinery today I
am going to take him seriously.
     It really boggles the mind to contemplate such a "general data mining
program" wending its way around the Internet, ferreting out knowledge here
and there with understanding to surpass that of a brilliant human. It
sounds science fiction-like. So do those Cisco tv ads which have been
running since 1999. "One day, training for every job on earth will be
available on the Internet." The Cisco ad implies that the trainees will be
human. If however, the data mining software is as advanced in a decade as
Kurzweil expects, could the trainees not be robots too?
     Those "business objectives of an enterprise" are now being met by
intelligent machinery which is acquiring knowledge, ie learning. When the
"peripherals" of our desk top computers expand to become complex receptors
like artificial vision and hearing systems and complex effectors like
robotic arms and grippers, why would the data mining programs not tell us
to use these devices instead of human labor to meet business objectives?
Consider even those "smart homes" which are now being built at least as
prototypes. If they are really smart they will have more than a computer
which regulates temperature and humidity. It should be able to take care
of most everyday chores. The data available on the web will be able to
tell the smart home how to get even smarter...automatically. And why
couldn't it do the same for a "smart office" or "smart factory"?
FWP.


------------------------------------------------------------------------
Special Offer-Earn 300 Points from MyPoints.com for trying @Backup
Get automatic protection and access to your important computer files.
Install today:
http://click.egroups.com/1/2344/5/_/433155/_/955390654/
------------------------------------------------------------------------

           *** The Era of Total Automation is Now ***

• Prev by Date: DM: DIMACS/IBM Workshop on Data Mining in the Internet Age (May 1-2, 2000)
• Next by Date: DM: Transforming Data
• Prev by thread: Re: DM: Transforming Data
• Next by thread: DM: DIMACS/IBM Workshop on Data Mining in the Internet Age (May 1-2, 2000)
• Index(es):
  • Date
  • Thread