This page last changed on Nov 24, 2010 by barbosaa.
Description

This algorithm generates a merging table as described in Create Merging Table . The merging table will be pre-annotated with merges based on the criteria selected. Specifically, the UNSPLIT_NAME column values will be compared using the given string similarity measure, and those above a certain threshold will be indicated to be part of the same merge group. A primary entity for the group will be selected based on name fields filled in and their length.

Usage Hints

You will want to check this table thoroughly. Merging identical names already has a chance of merging names that are not really the same person; merging based on string similarity nearly guarantees some names that are really different people will be marked the same. After separating those merge groups, use Merge Entities to perform the merge.

Document generated by Confluence on May 31, 2011 16:37