The labeler assets that you create are available to the users who create mappings in Data Integration. You or other users add a Labeler transformation to a mapping and add a labeler asset to the transformation. The Labeler transformation applies the logic in the asset to the data source that the mapping identifies.
A Labeler transformation works in a similar way to a Mapplet transformation. A Data Integration user connects the transformation inputs and outputs to other objects in the mapping in the same manner as a mapplet connects to other mapping objects. When the Data Integration user runs the mapping, the Labeler transformation applies the labeler asset logic to an input field and generates output data for downstream objects.
The asset logic comprises the labeling criteria that you define in one or more the steps in the asset. The transformation applies the asset logic to a single input field.
Token labeling output fields
When the mapping runs, the Labeler transformation creates the following output fields for token labeling:
LabeledOutput
A copy of the input field data in which any value that matches the logic in an asset step is replaced with the label that the step specifies.
TokenizedData
A copy of the input field data. If the asset includes a dictionary step, you can optionally configure the step to replace any value that matches a dictionary value with the corresponding valid value from the dictionary.
You can test the run-time performance of a labeler asset in Data Quality. The asset writes test results to the Labeled Output field and the Tokenized Data field.
Character labeling output fields
When the mapping runs, the Labeler transformation creates the following output fields for character labeling:
LabeledOutput
A copy of the input field data in which any character that matches the logic in an asset step is replaced with the label that the step specifies.
You can optionally configure a step to ignore any input term that you specify.
If the asset includes a dictionary step, you can optionally configure the step to override any later step in the sequence.
Input
Contains the input data.
You can test the run-time performance of a labeler asset in Data Quality. The asset writes test results to the Labeled Output field.