AnsweredAssumed Answered

Duplicate Records Szenario

Question asked by CloudMeister on Aug 15, 2011
Latest reply on Aug 16, 2011 by CloudMeister


Duplicate Records Szenario


Hi all,

since I am still relatively new to FileMaker I would like to see if any of you have a better idea for how to handle the following Scenario:

The Problem:

Customer Database is without table for customers, but with some 24 000 Orders incl. Customer Address information.

Wild typographical errors and styles (examble East Northbrook drive vs. E. Northbrook Dr.)

The Vision:

Obtain Unique Customers out of that Database and relate them from a customer table to the orders table

My Solution:

Using the 360 Works Scriptmaster Plugin Module TextSimilarity we can calculate how similar records are. We are comparing the following fields: Billing Company Name, Billing Street Address & Billing ZIP Code

In order to compare 24 000 Records to eachother I have written a script that does the following:

1) Set a variable for a draft cust ID (this variable is counted up as used)

2) Loops through all Records that do not have a draft cust ID Assigned in order to assign the same draft cust ID to similar records. Individual records without similarity to other records have their own individual cust ID.

The script is running as I write this. Eventhough it is running locally on one computer it looks like it will be running 3+ Days.

After that a human will still have to go through the individual Records in order to see if there are errors and to make corrections as needed.

As this is an extremely timeconsuming method I am looking for something a little more performant. Any Ideas?


Regards Ernst

     - find close matches with