Standardization workflow

When you standardize data, the workflow depends on your data cleansing goals, data domain, and experience with pattern-action language.

You can prepare for the standardization process in the following ways:

Ensure that you understand the data quality requirements for the domain. If a subject matter expert for the domain is not available, you might need to conduct research about the domain.
Prepare a representative sample data set.
Analyze the source data by running a job that includes the Investigate stage.

The following diagram shows a workflow for the standardization process. Click a section of the diagram to view a topic about that part of the standardization workflow.

The diagram shows a standard workflow for the standardization
process. You get domain-specific data to standardize, then determine
whether a predefined rule set applies to the domain. If a predefined
rule set applies to the domain, you provision the rule set, apply
the rule set to your data in a job, and then use an Investigation
or SQA report to assess the output data. If the standardized data
meets your data quality goals, the standardization process ic complete.
If a predefined rule set does not apply to your domain, you can create
a new rule set or copy and customize a predefined rule set. If you
are experienced with pattern-action language, you can modify the pattern-action
file directly. If you are not experienced with pattern-action language,
you can enhance the rule set in the Standardization Rules Designer.
After you enhance the rule set, you provision the rule set, apply
the rule set to your data in a job, and then use an Investigation
or SQA report to assess the output data. If, at any time after you
assess the output data, the standardized data does not meet your data
quality goals, you can continue to modify or enhance the rule set.