How do I resolve duplicates?

Duplicate resolution is a set of processes that you can use, after duplicate records are identified, to consolidate those records.

You can resolve duplicates in two ways, either by linking them or by merging them. Linking involves associating the duplicate records.

The linked records are treated as unique records in the data registry, and have their own unique identifiers. Merging involves combining duplicate records into one new master record.

You can setup linking by configuring a couple of profile options discussed in the topic Duplicate Resolution Simplified Profile Options. However, the setup for merging is a bit more elaborate. It consists of configuring logic to:

• Determine which record from a set of identified duplicates should be designated as the master record. You can set this up by configuring Set Master Record Rules.
• Determine which attribute value instances from across the set of duplicates the master record should contain. You can set this up by configuring Set Attribute Value Rules. The Set Master Record Rules and Set Attribute Value Rules are collectively called Survivorship Rules.
• Determine whether the merge is violating any of the conditions under which a merge should be prohibited. You can set this up by configuring Agreement Rules.

You can use either use Groovy Scripts or Oracle Business Rules to define these configurations.

In case of Accounts, Customer Profile Quality Scoring is an additional method to resolve duplicates. This method helps in assigning scores to records and identifying a record with the highest score. This record with the highest score can be used as the single account record, instead of multiple duplicate records.