Re: CMJ 28:4


Subject: Re: CMJ 28:4
From: Eliot Handelman (eliot@generation.net)
Date: Sat Feb 05 2005 - 16:02:45 EST


Doug Keislar wrote:

>
> Music Classification
>
> Algorithmic Clustering of Music Based on String Compression
> Rudi Cilibrasi, Paul Vitányi and Ronald de Wolf

interestingly I read this article last week, and I think I mentioned it
on my show.
The idea -- kolmogoriv compression -- is interesting, but the authors
admit that
their results were oin thje order of the dog who walks on his hind legs,
ie, not well.
Currently they're working on application of same algoithm to create
autoimatic semantics
by categorizing pairs of words through google, and got slash-dotted last
week.
Interesting, but not worthy of publication in a music specializing
journal, since their algorithms
reveal nothing about music.

-- eliot



This archive was generated by hypermail 2b27 : Sat Dec 22 2007 - 01:46:06 EST