From: Craig Henderson (cdm.henderson_at_[hidden])
Date: 2002-10-21 02:22:17
"Bohdan" <warever_at_[hidden]> wrote in message
> I was looking for a theory on such kind of algorithm some
> time ago, but first ~20 links from google gave me only definition of
> no algorithms :(. So i could not prove that my implementation really gives
> count of operations to transform string. The next hole is that i do not
> if there is more effective algorithm than my ... no doubts it exists :).
> We used string difference in UDF (user defined library) on SQL server
> namely for comparing person address considering possible errors.
> Speed requirements for such function were really high.
> I'd like to look at your/Levenstain/Lalande implementaion and
> i will appreciate if you give me few links.
try searching for "levenshtein distance" in Google, rather than "edit
distance". The site I used most for my reference during developing the
algorithm was http://www.merriampark.com/ld.htm This has a java applet
aswell, so you can compare the results.
I'll put the edit distance algo in the sandbox asap, but not untill Tuesday
evening UK time at the earliest, I'm afraid
Boost list run by bdawes at acm.org, gregod at cs.rpi.edu, cpdaniel at pacbell.net, john at johnmaddock.co.uk