Standardizing Values with a Standardization Scheme

Identify Incorrect Values

You can use DataFlux Data Management Studio to get detailed information about problems with source data. For example, you could identify all of the incorrect spellings of a company name in a table column. Detailed information about incorrect values will help you create an effective standardization scheme. For more information, see Analyzing the Quality of Data Sources.

Create a Standardization Scheme

Use DataFlux Data Management Studio or the DQMATCH procedure in SAS Data Quality Server to create a standardization scheme that will map incorrect values to the correct ones. The next figure shows a scheme in DataFlux Data Management Studio that can be used to correct misspellings for the Computer Furniture company.

Standardization Scheme for a Company Name

For more information about creating standardization schemes, see the scheme topics in the “Customize” chapter of the DataFlux Data Management Studio User’s Guide. Alternatively, see the documentation for the DQMATCH procedure in the documentation for SAS Data Quality Server.

Verify Prerequisites

The Apply Lookup Standardization transformation that is used in this topic requires SAS Data Quality Server 9.3. In SAS Data Integration Studio, verify that the appropriate Scheme Repository Type and Scheme Repository are selected, as described in Global Options on the Data Quality Tab. The scheme repository must contain the standardization schemes that you want to use in SAS Data Integration Studio.

Note: On the Data Quality tab, if you change an existing value in the fields Scheme Repository Type or Scheme Repository, then you must replace any instances of the Apply Lookup Standardization transformation in any existing jobs that you intend to run using your current metadata profile. Replacement is required because scheme metadata is added to these jobs when they are run for the first time. To update a job to use a different scheme repository, add a new Apply Lookup Standardization transformation to the job, configure the new transformation, delete the old transformation, and move the new transformation into place.

Create and Populate the Job

The example job that is described in this section uses an Apply Lookup Standardization transformation. This transformation applies one or more standardization schemes to one or more columns in a source table. Applying schemes modifies your source data according to rules that are defined in the schemes. The specific process of scheme application varies based on your input. However, in general, when you apply a scheme to a source column, each value in that column is compared to all data values in the scheme. If the source value matches a scheme data value, the associated standardization value in the scheme is written into the target as a replacement for the source value. If no match is found, the source value is written into the target without change.

The first task is to create a job flow that reads a table with nonstandard data (MANUFACTURERS), uses a standardization scheme to correct the data, and then writes the corrected output to a target table (MANUFACTURERS_STANDARDIZED). The flow would look similar to the following figure:

Example Job with an Apply Lookup Standardization Transformation

Perform the following steps to create and populate the job.

Create an empty SAS Data Integration Studio job.
In the Data folder of the Transformations tree, drag the Apply Lookup Standardization transformation into the empty job in the Diagram tab.
Select and drag a source table from its folder and drop it before the Apply Lookup Standardization transformation. In this sample job, the name of the source is MANUFACTURERS. The source provides contact information for suppliers of computer equipment. In the MANUFACTURERS table, the Name column contains inconsistent values for the supplier named Computer Furniture, as depicted in the following display:

Source Table Data with Errors in the Name Column
Drag the cursor from the source table to the input port of the Apply Lookup Standardization transformation. This action connects the source to the transformation.
In the Access folder of the Transformations tree, drag the Table Loader transformation into the empty job in the Diagram tab.
Drag the cursor from the output of the Apply Lookup Standardization to the input port of the Table Loader transformation. This action connects the two transformations. .
Drag the target table from its folder and drop it after the Table Loader transformation on the Diagram tab. In this sample job, the name of the target is MANUFACTURERS_STANDARDIZED. The target has the same columns as the source.
Drag the cursor from the output port of the Table Loader transformation to the target table. This action connects the transformation to the target. The job flow should now look similar to Example Job with an Apply Lookup Standardization Transformation.

Configure the Apply Lookup Standardization Transformation

The goal for this task is to associate the standardization scheme to the column or columns in the source that contain inconsistent values. This is done by selecting options on the Standardizations tab in the Apply Lookup transformation. An example set of options is shown in the next figure.

Options Selected on the Standardizations Tab

Perform the following steps to configure the Apply Lookup Standardization transformation:

Open the properties window of the Apply Lookup Standardization transformation and display the Standardizations tab.
Right-click the down arrow in the Locale field to display the available locales. Select the locale that represents the national language and region that best represents your data. In the sample job, you could select ENUSA (English language, as implemented in the United States of America).
Specify the schemes to be applied to specified columns. In the sample job, right-click in the table cell of the Name row and the Scheme column. This action displays a list of available schemes in the scheme repository.
Select the scheme to be applied to the column. For the sample job, this is a scheme named Manufacturer_Names, which was created as described in Create a Standardization Scheme..
Click the Apply Mode column and select Phrase, which applies the standardizations to the entirety of each character string in the Name column.
The next step is to specify a value in the Lookup Method column. If you accept the default value of Exact, then only an exact match in your scheme will result in a corrected value being written to the target table. Alternatively, you could use match definitions as described in steps 7–9.
(Optional step) If appropriate match definitions are available in the selected locale, you could click the Lookup Method column and select Use Match Definition. Selecting Use Match Definition activates two related fields.
(Optional step associated with match codes) Click the Definition column to display a list of available match definitions. A match definition aims to help you decide whether two or more pieces of data might refer to the same real-life entity. To facilitate this, the definition generates a special string called a match code for each input. Any two inputs that generate the same match code are considered a match. Select a definition that is appropriate for the current column.
(Optional step associated with match codes) Use the Sensitivity column to control the precision of the match. A lower number is a less-exact match.
Click OK to save your input and close the properties window. The job is now ready to be run.

Run the Job and View the Output

Perform the following steps to run the job and view the output:

Right-click on an empty area of the job, and click Run in the pop-up menu.
After the completion of the job, right-click the target and select Open to view the standardized contents of the Name column. Note that one source value (Comp Furn) was not mapped in the standardization scheme that was created in Create a Standardization Scheme. All the other values were standardized. The following figure shows the target table data for the sample job.

Standardized Name Column in the Sample Target Table