The Tune rules section attempts to improve your current ruleset by training a machine learning algorithm with user provided match outcomes. The algorithm makes continuous changes to your ruleset, scoring the ruleset after each change to determine which set of changes result in duplicate matches most closely matching the user provided match outcomes. As Tune rules runs, you will be able to see how the current modified ruleset performs against your original ruleset.
To begin, press the Start button. You will be presented with two records from your duplicate store - using the Yes, Maybe, and No buttons, select whether the two records should be considered a duplicate match. You will continue to be presented with new pairs of records. The rules tuning will begin once you have provided enough expected match outcomes - once this happens, you will be able to see the rules tuning status at the bottom of the page. The more expected match outcomes you provide, the more data the machine learning algorithm will have to produce a better ruleset.
Once Tune rules is running, the bottom of the page will display two tables - one showing how well the default ruleset performs and the other showing how well the tuned ruleset performs. To understand how well each ruleset is performing, compare the inferred match result (the result returned by the ruleset) against the actual match result (the result you provided). A higher overlap between the inferred and actual match result means a better performing ruleset.
Clicking on a value in the results table will let you view all the input record pairs that make up that value. If you want to change the expected match result for any pair, you can click either of the records for that pair. This will place the pair of records back into the scoring section so you can use the Yes, Maybe or No buttons to score it again.
To download the ruleset generated by Tune rules, click the Download Ruleset button. You can then explore the modified ruleset using Compare records or the Ruleset editor by clicking on Change settings and replacing the rules with the modified ruleset. If you are happy with the modified ruleset, you can then use it in your Find duplicates step configuration.
