Rick Jelliffe writes:
> The soundex and multiphone algorithms convert strings to sound
> equivalents: a kind of multiphone algorithm is probably similar to
> what you are looking for. These were created to allow hashed
> lookup up family names based on sounds, and (in the case of soundex
> at least) work on the assumption that spelling mistakes are more
> common later in words than earlier.
Thank you. I am familiar with soundex, but unlike a hashing
algorithm, it does not produce a fixed-length key -- all other things
(proportion of vowels and significant clusters) being the same,
a longer string will have a longer key.
All the best,
David
--
David Megginson, , http://www.megginson.com/
