Path: utzoo!utgpu!news-server.csri.toronto.edu!clyde.concordia.ca!uunet!cs.utexas.edu!rutgers!mcnc!unccvax!hjingyi
From: hjingyi@unccvax.uncc.edu (Hu Jingyi)
Newsgroups: comp.ai.neural-nets
Subject: help!
Message-ID: <2428@unccvax.uncc.edu>
Date: 7 Jul 90 21:43:26 GMT
Organization: University of NC at Charlotte
Lines: 19


Dear Sir/Madam :

I have a question about the Levenshtein distance. in "Self-Organization and
Associative Memory" book (page 67, 4 lines) T.Kohonen written:

weighted Levenshtein distance  WLD(A,B) = min{pa(i) + qb(i) + rc(i)}   (2.93)

where the coefficients p, q, and r may be obtained from the so-called
confusion matrix of the alphabet, as the inverse probability for particular
type of error to occur.

Could you help me to get the p, q, and r confusion matrix of the alphabet?
I want use the distance for some words matching processing application.

Thanks a lot.

Regard
Hu Jingyi