1.3.4.9.17 Match Transformation: Replace

The Replace transformation allows you to use a Reference Data Map in order to standardize data for the purpose of either clustering or matching.

This transformation works in exactly the same way as the main Replace processor in EDQ. This help page describes the typical use of the Replace transformation when matching.

The Replace transformation is very useful to overcome the variations of the same value that users often enter into free text fields.

There are many cases where the same value may be represented in a number of different ways. For example "Bill", "William", and "Billy" are all different forms of the same first name, and "Hse", "House" and "Ho." are all forms of the word "House". A Reference Data Map allows such synonyms to be standardized to a single way of representing the value.

For example, the following entries could exist in a Reference Data Map, and be used to standardize first names for the purposes of matching:

Table 1-89 Reference Data Map Values

Value	Standardization
Bill	William
Billy	William
Willy	William
Mike	Michael
Mick	Michael
Mickey	Michael
Dave	David
Jim	James

The following table describes the configuration options:

Configuration

Description

Options

Specify the following options: Reference data: matches the attribute values against the lookup column in the map. Where there is a match, the matching value is replaced by the value in the right-hand column. Type: Reference Data. Default value: None. Match longest value first?: controls which replacement to perform where there are multiple matches against the map, in Starts With, Ends With, or Contains replacement. Type: Yes/No. Default value: No. Ignore case?: determines whether or not to ignore case when matching the lookup column of the map. Type: Yes/No. Default value: Yes. Match by: drives how to match the map, and therefore which part of the original value to replace. Type: Selection (Whole Value/Starts With/Ends With/Contains/Delimiter Match). Default value: Whole Value. Delimiters: when matching values to the map by splitting the data using delimiters, this allows you to specify the delimiter characters to use. Type: Free text entry. Default value: Space.

Example

In this example, the Replace transformation is used to standardize first names for the purpose of matching.

Example configuration

Reference Data:

Table 1-90 Example Reference Data

Value	Map	Active
Bill	William	Yes
Billy	William	Yes
Willy	William	Yes
Will	William	Yes
Mike	Michael	Yes
Micheal	Michael	No
Mickey	Michael	Yes
Dave	David	Yes
Steven	Stephen	Yes
Steve	Stephen	Yes
Jim	James	Yes

Match longest value first?: No

Ignore case?: Yes

Match by: Whole Value

Delimiters: <space>

Example transformations

The following table shows example Replace transformations using the above configuration:

Table 1-91 Example Transformations for Replace

Value	Transformed Value
Steven Lewis	Stephen Lewis
Stephen Lewis	Stephen Lewis
David Stevens	David Stevens
David Steven	David Stephen
Mike Davis	Michael Davis
Micheal Lewis	Micheal Lewis
Mickey Lewis	Michael Lewis
Jim Jones	James Jones
James Jones	James Jones
Bill Taylor	William Taylor
Will Taylor	William Taylor