The Relationship Linker process identifies the duplicates in the same entity. It organizes records sharing the same window keys into match windows and compare records within the match window. It then generates a statistics file summarizing the matching results and appends the matching information to the matched records in the output.
Before configuring and running the Relationship Linker process, you need the following preparation:
- Create window keys using the Window Key Generator. The window keys must be produced and included in the input entity.
- Sort rows by the window key using the Sort Utility. After creating the window keys, but before running the Relationship Linker, the input record must be sorted by the window key.
Relationship Linker Process
The Relationship Linker completes its process in three major steps:
Step 1: Create match windows
The Relationship Linker uses the window keys to generate the match window.
Step 2: Compare records
All records within the match window are compared to each other. Matches are determined based on the matching attributes, comparison routine, score, grade, and pattern. Based on the grade pattern returned by the routines, the result of the comparisons is considered a Pass (matched), a Suspect (suspect match), or a Fail (unmatched). See Determining Matches for details.
Step 3: Generate results
Records are then sorted so that matched record sets are grouped together and appear first in the output. The Relationship Linker will generate a statistics file summarizing the matching results and an output appending the various matching information. For example, it will provide Pass and Suspect records with the same match key (lev1_matched, lev1_suspect, etc) identifying whether the record is Pass or Suspect. See Viewing Relationship Linker Results.
The process ends when the match window is cleared. Window matching is repeated for each window key set in the input entity until there is no more matching windows left.
Click any of the following topics to learn more about how to setup and run the Relationship Linker process.
- Determining Matches
- Schema Editor
- Input and Output Settings
- Process Settings
- Editing Advanced Rules
- Running the Relationship Linker process
- Viewing Relationship Linker Results
- Tuning the Linking Rules
By default, Name and Address projects include a Relationship Linker process. If your business goals require reference matching, you can convert Relationship Linkers to Reference Matchers. You can also convert Reference Matchers to Relationship Linkers.
When you deploy a project for real time, the deployment process automatically prepares the project to perform either a reference match or a window match. See the topic Deploying a Real-Time Project for more information.