Processing Duplicates: Four Steps to Follow

 

The process you would undertake to examine each pair of matched records would be as follows:

 

STEP 1: Determine if it’s a true duplicate: You would first check the match to see if the system has mistakenly identified a match (called a False Duplicate).  If the two identified duplicates are clearly not the same record (i.e. not duplicates after all) then click the Ignore link to eliminate the match and move onto the next one.

 

STEP 2: Select the record you want to keep: This involves selecting the record that contains the most complete set of data and making sure that record is the Master record (the record on the left – will be kept during a merge).  If this record is on the right as a slave record (the one that will be deleted during a merge) then you want to click on its primary key link (this is directly under the likeness score).  This will swap the records over, moving that record into the master position and the master record into the slave position.

 

STEP 3: Make any manual edits and copy-over’s: Examine any data conflicts found (highlighted in red cells) and determine which cells in the master record need to be edited to ensure data is not lost when the slave record is deleted.  You can copy the entire contents of a slave cell over to the master by pressing the  Copy-Over button.

 

STEP 4: Merge the records together: Click the Merge link to complete any auto-merging of fields into the master record, and then delete the slave record.  You will then be moved onto the next record ready for processing.

 

For an example of these steps in action see Processing a Duplicate Match: An Example

Once all the duplicate matches have been processed you will need to Export the Data back into your live systems.

 

Related Topics

Processing Tool