Do you mean an algorithm that compares the standard *written* representation of words and decides if they are homophones, like ruff : rough or write : right?
 
What particular languages are you working with?
 
Piotr
 
 
----- Original Message -----
From: stefan.nonneman@...
To: phoNet@egroups.com
Sent: Sunday, January 07, 2001 7:24 PM
Subject: [phoNet] Phonetic Algorithms !

High,

I am looking for phonetic algorithms able to detect words with
almost the same sounds. I found an algorithm in PERL working fine
in the English context. I want to apply this algorithm in order
to find typing errors in databases. Working with databases
in 11 languages, I need the same algorithm but also for the 10 other
languages.

Can someone help me?

Also suggestions on how to detect +/- identical words in one
language are very welcome.

Best regards,

Stefan