The documentation portal is undergoing maintenance. During this period, you can't create any new user credentials. If you do, you'll need to recreate the user credential after the maintenance period ends.
Duplicate Record Exception Transformation Overview
Duplicate Record Exception Transformation Overview
The Duplicate Record Exception transformation is an active transformation that reads the output of a data quality process and identifies duplicate records that require manual review. The Duplicate Record Exception transformation is a multiple-group transformation.
The Duplicate Record Exception transformation receives input from another transformation or from a data object in another mapping. The input to the Exception transformation must contain a numeric match score that the transformation can use to determine if the record is a duplicate. Set an upper and lower match score threshold in the Duplicate Record Exception transformation.
The Duplicate Record Exception transformation performs one of the following actions:
If the match score is greater than or equal to the upper threshold, the transformation treats the record as a duplicate and writes it to a database target.
If the match score is less than the upper threshold and greater than the lower threshold the transformation treats the record as a possible duplicate and writes it to a the record to a different target for manual review. If the record belongs to a cluster, the transformation writes all records in the cluster to the target.
When a cluster has any match score less than the lower threshold, all records in the cluster go to the unique records output group. Clusters of size 1 are routed to the unique group, regardless of match score. By default, the Exception transformation does not write unique records to a target. You can configure the transformation to return the unique records.
If any match score in a cluster is not in the range 0 - 100, the Exception transformation ignores all rows in the cluster. The Data Integration Service logs a message that includes the clusterID.