Home » RDBMS Server » Server Administration » Duplicates users - fuzzy!
Duplicates users - fuzzy! [message #373749] Fri, 04 May 2001 04:37 Go to next message
Robert Angel
Messages: 8
Registered: March 2001
Junior Member
Does anyone know how I can modify the below to give me nearly the same as (i.e. a fuzzy search for duplicates)

Select description, count(user_id)
from fnd_user
having count(user_id) > 1
group by
description;

n.b. description is the thing containing users name upon which I want to pursue the fuzzy search, user_id being the prime key.

Thanks for any light on this,
Robert

p.s. can the sensitivity of the fuzziness be 'adjusted' to manage the number of hits returned?
Re: Duplicates users - fuzzy! [message #373755 is a reply to message #373749] Fri, 04 May 2001 13:39 Go to previous messageGo to next message
Sundar Venkatasubramaniam
Messages: 26
Registered: May 2001
Junior Member
can you be more specific with examples so that it will be helpful for any suggestion.
Thanks
Re: Duplicates users - fuzzy! [message #373760 is a reply to message #373749] Fri, 04 May 2001 17:06 Go to previous messageGo to next message
Andrew again...
Messages: 270
Registered: July 2000
Senior Member
Run this example

create table fuzzy (col1 varchar2(20))

insert into fuzzy values ('localized')
insert into fuzzy values (' localized ')
insert into fuzzy values ('localized ')
insert into fuzzy values ('localised')
insert into fuzzy values ('Localized')
insert into fuzzy values ('macdonalds')
insert into fuzzy values ('mcdonalds')
insert into fuzzy values ('mac donalds')
insert into fuzzy values ('mc donald''s')
insert into fuzzy values ('hello world')

select upper(trim(col1)), count(*)
from fuzzy
group by upper(trim(col1))

select soundex(upper(trim(col1))), count(*)
from fuzzy
group by soundex(upper(trim(col1)))

select col1||'<--' from fuzzy where soundex(upper(trim(col1))) = 'L242'
Re: Duplicates users - fuzzy! [message #373778 is a reply to message #373755] Tue, 08 May 2001 05:49 Go to previous messageGo to next message
Robert Angel
Messages: 8
Registered: March 2001
Junior Member
what I had in mind is if a user had two ids, and, say, they put their name as Richard Grant and on the other log-in Richard E Grant - purchases, what algorithm could I employ to highlight this as a 'probable match'
Re: Duplicates users - fuzzy! [message #373779 is a reply to message #373760] Tue, 08 May 2001 05:59 Go to previous messageGo to next message
Robert Angel
Messages: 8
Registered: March 2001
Junior Member
Thank you for that!
Could you explain the significance of 'L242' in the example and how I can vary the 'sensitivity' of its matching?
Re: Duplicates users - fuzzy! [message #373780 is a reply to message #373760] Tue, 08 May 2001 05:59 Go to previous message
Robert Angel
Messages: 8
Registered: March 2001
Junior Member
Thank you for that!
Could you explain the significance of 'L242' in the example and how I can vary the 'sensitivity' of its matching?
Previous Topic: SELECT statements work, INSERT doesn't.
Next Topic: Interrupting a query
Goto Forum:
  


Current Time: Mon Dec 23 10:20:45 CST 2024