Need query to find repeating values, please help

Question

I need help with SQL or whatever you guys think is the best way to do this, maybe PL/SQL.
Basically it like this, in the database 10G, there are two field for address:
ADDESS_DELVRY - street address for example would be stored here
ADDESS_DELVRY2 - apartment building, suite number or PBM would be stored here
but sometimes I see the Apartment B-123 in both fields, so I need to find those duplicates, if there is a way.
So for example:
ADDESS_DELVRY - 123 Elm Street Apartment 2B
ADDESS_DELVRY2 - Apartment 2B
This is a problem, because then the Apartment gets printed twice on the mail piece. I think there is a small percentage that have this problem and am trying to find out the best way to ID them, maybe use REG_EXP ??
Not sure, please help, thanks!

Answer 1

Soundex works how it's designed to work, which unfortunately is a very poor approximation at words sounding the same.
Using the UTL_MATCH package is a far better way to check for similarity between strings.
Ok. Given the OPs challenge here; a simple, or messy (we don't know), data cleaning task, I think soundex was worth mentioning.
Also, the reply that "soundex don't [doesn't] work" , given a working example , I think we can safely conclude
that UTL_MATCH is way above his league.
I don't understand how the OP has better use of
SELECT UTL_MATCH.EDIT_DISTANCE('Second Avenue 12 second floor', 'second fl.') dist FROM DUAL;
      DIST
        20
1 row selected.than
SELECT *
  FROM dual
WHERE SOUNDEX ('Second Avenue 12 second floor') = SOUNDEX ('second fl.')
X
1 row selected.given the task at hand, but all right I'm not going to argue either.