Match Cleansers

The cleansers provided by Reltio enchance the matching process.

Reltio provides three out-of-the-box data transformation capabilities that you can leverage exclusively within the execution of a match rule to increase the effectiveness of matching. They are:

  • Name Dictionary Cleanser activates and controls synonym matching
  • String Replacement Cleanser can find a predefined substring in an attribute and replace it with another string of your choice.
  • “Noise words” removal capability can eliminate unwanted noise words from an attribute during matching.

All three capabilities perform their function only within the match framework and do not modify the actual data in the record. (These services should not be confused with the platform’s profile-level cleansers (For example: email, phone, address cleansers) which actually do modify the profile data and are documented here. For more information, see Out-of-the-box Cleanse Functions.

To be clear, the first two items (Name Dictionary Cleanser, String Replacement Cleanser) are stand-alone cleansers you can enable using the Cleanse element within the match rule. Conversely, the noise words removal capability has to be enabled through a comparator and/or match token class. Two of the comparator classes and two of the match token classes offer in-built noise words removal using internal lists of words developed by Reltio. But you can leverage a list of your own design by creating a custom comparator and custom match token class and declaring your list with that custom class.

Order of execution

When a cleanser is present in a match rule, it executes before the match token class and before the comparator class. It effectively replaces the attribute’s values with a cleansed (aka “transformed”) set of values that the match token and the comparator class will use for input.