Re: N-Gram

[collins@newton.apple.com (Scott Collins)]

  >[algorithm to] store data the same way the human brain does.
  >[stored data would] take up only [0.5%] of the original space.

Whoever said the human brain stores data compressed to 0.5% of its original
size, and what is its original size anyway.

Paul Baclace says:
  >Sounds like Bugajsky creates a generative grammar and then stores list
  >of productions that specifies a walk on the tree to extract data.  This
  >is a form of Kolmogorov Complexity compression, which has been expanded
  >upon most notably by Chaitin.

I agree.  The description sounds more like this than anything else I'm
familiar with.

Paul Baclace goes on to say:
  >I wonder whether [he] includes the size of his grammar in [the claim]

0.5% is a questionable claim.  If it includes the grammar, then the grammar
must be very simple, and the data of very low entropy with respect to it --
in which case 0.5% would be an uninteresting experimental result.  If the
claim does _not_ include the size of the grammar, then the claim is useless
for evaluating this scheme.


Scott Collins         | "Few people realize what tremendous power there
                      |  is in one of these things."     -- Willy Wonka
......................|................................................
BUSINESS.   voice:408.862.0540  fax:974.6094   collins@newton.apple.com
Apple Computer, Inc.   5 Infinite Loop, MS 305-2B   Cupertino, CA 95014
.......................................................................
PERSONAL.   voice/fax:408.257.1746    1024:669687   catalyst@netcom.com