Hi All,

My app is supposed to find duplicate contacts of an user. The logic is to find a duplicate contact on the basis of email, name, and mobile number. If there is a match found in either of these attributes of a contact, then this contact can be considered as duplicate of that contact with which match was found.

I would like the group to guide me on the best possible design to handle this: I can think of following but please suggest which option can give me best performance:

1) Simply query all the contacts and then on each contact, do an iteration to find the matches
2) Make use of Hibernate search and index name, email and mobile
3) Make use of a Stored Procedure and then invoke it from Hibernate. SP will find all the duplicates.
4) Any other approach

rgds
nitin

It depends of your definition of "duplicate", if you need to remove automatically all exact matches, a SP would perform best.
If you need some more flexibility I'd go for Hibernate Search, as you could scroll on all your users and for each one verify from the index if there are other users "very similar" to the current one, where you can define the "similarity" function in a flexible way (like ignoring small typos in address fields, telephone numbers which are the same but formatted differently, names having or not having titles in it...)