Siebel Data Quality Administration Guide > Using Siebel Data Quality > Merging of Duplicate Records >

Sequenced Merges

You use a sequenced merge to merge multiple records into one record. You assign sequence numbers to the records so that the record with the lowest sequence number becomes the surviving record, and the other records, the nonsurviving records, are merged with the surviving record.

When records are merged using a sequence merge, the following rules apply:

  • All non-NULL fields from the surviving record are kept.

    Any fields that were NULL in the surviving record are populated by information (if any) from the nonsurviving records. Missing fields in the surviving record are populated in ascending sequence number order from corresponding fields in the nonsurviving records.

  • The children and grandchildren (for example, activities, orders, assets, service requests, and so on) of the nonsurviving records are merged by associating them to the surviving record.

Sequenced merge is especially useful if many fields are empty, such as when a contact record with a Sequence of 2 has a value for Email address, but its Work Phone # field is empty, and a contact record with a Sequence number of 3 has a value of Work Phone #. If the field Email address and Work Phone # in the surviving record (sequence number 1) are empty, the value of Email address is taken from the records with sequence number 2, and the value of Work Phone # is taken from the record of sequence number 3.

A sequence number is required for each record even if there are only two records.

Field Characteristics for Sequenced Merges

A field must have specific characteristics to be eligible for use in a sequenced merge:

  • The field can not be a calculated field and must reside on a physical database column.
  • The field must be active, that is designated as Active in the respective business component.
Siebel Data Quality Administration Guide Copyright © 2006, Oracle. All rights reserved.