Duplicate Query

[risoam]

I have a table of names and addresses where for each record, there may be 3 more records of the same person just spelled a little differently. For instance, record 1 may be John Smith at 120 Tree Rd. Record 2 may be J Smith at 120 Tree Rd. Record 3 may be John Smith at 120 Tree Road. I am looking for a way to find all of these matching records and make one table with 1 record for each person. So when there was John Smith 3 times my output table would only have one instance of John Smith.

 

[blom0344]

hello risoam,

Your example is a typical form of data-contamination that is best solved by hiring someone to do the dirty work by hand.

SQL is not suited to tackle this sort of problem, you would need to start some sort of neural network to solve this and that would be even more expensive.

Main reason why SQL would be letting you down is that you still would be forced to foresee every possible scenario and create the corresponding code to deal with it. Foreseing every possible scenario would mean going through all your data.

By the time you have you this figured out (let alone build) you could have made the adjustments by hand.

sorry