[Pdx-pm] Encoding for fuzzy matching

Roderick A. Anderson raanders at acm.org
Thu May 11 08:12:36 PDT 2006


David Wheeler wrote:
> On May 9, 2006, at 15:03, Rafael Almeria wrote:
> 
>> Soundex may allow you to do a very fuzzy matching for text.
> 
> 
> And Text::LevenshteinXS is even better.

Looked promising but then fell through for what I need to do.  I'm 
beginning to think I'll have to CASS certify(/correct) the addresses and 
then soundex ( or some other system ) encode each portion of the 
address: number, [per-directional,] street, [post-drectional,] city, 
state [, zip_postal, ] [country].

And writing this makes me think I have it all buggered up.  Sort the 
list from the database, and look for those that are similar to the 
previous record.

Thanks to all for the suggestions.


Rod
-- 


More information about the Pdx-pm-list mailing list