When the Unified Data Model (UDM) identifies a record as a potential match (i.e., a possible duplicate of an existing record), it is flagged for manual review. During this review, a data steward or administrator may determine that the flagged record is not actually a duplicate or should not be merged with the existing record. In such cases, the steward can choose to reject the record from the potential matches list.
Why Would You Reject a Record in Potential Matches?
False Positives: Sometimes, the match rules (exact or fuzzy) may flag records as potential duplicates even though they represent different real-world entities (e.g., two customers with similar names but different addresses).
Business Context: The user may have additional business knowledge or context that the automated system does not, allowing them to make a more informed decision.
Data Integrity: Rejecting incorrect matches ensures that unique records are not mistakenly merged, preserving the accuracy and integrity of the golden repository .
The Process of Rejecting a Record in Potential Matches
Manual Review:
The user reviews the details of both the incoming (potential match) and existing records.
Decision to Reject:
If the user determines that the records are not true duplicates, they select the Reject action for the flagged record.