in reply to Re: Duplicate detection (SQL) in thread Duplicate (similarity) detection (SQL)
I assume to check on database level is the best solution but it is not working for everybody, e.g. if the database is popultated already. In that case you have to normalize the data (eg. the given adress: 24 thompsonrd., 10-03 is going to be normalized to -> thompsonroad 24, level 24, unitnumber 3 ...) and eventually validate the data. To do that properly you can easily spend a few hours :)
Then it becomes quite easy to check for duplicates:
data{"normalzed key"} =+ 1
Cheers
Hartwig
|