Document Understanding API guide

Validate the data by using the documentId and the classifierId or the extractorId. By calling this API, a validation task is created in Action Center.

Each action requires one POST and multiple GET calls (due to polling), meaning that if you want to validate a classification result, then you need to do a POST and multiple GET calls. Same applies for validating an extraction result.

When a POST call is initiated you need to add the classificationResults / extractionResults previously received when calling the classifier/extractor in the Request body of the POST request.

You can also pass a taxonomy as an input parameter to customize the validation experience. For example, you can change field hotkeys or the display color of fields.