Test learning instances: Standard Forms

Standard Forms are fixed format document types that have a defined template, with little or no variation in the overall layout, and are typically official or legal forms, such as a W-4 (Employee Withholding Certificate) form, income tax form, or a sale deed.

Prerequisites

  • Ensure your Control Room has the Document Automation (Number of pages) product license. For more details on licenses, see Understanding licenses in Document Automation.
  • You must have the AAE_IQ Bot Admin or AAE_IQ Bot Services roles to complete this task.
  • Ensure that you have identified between 5 (minimum) through 15 documents to train the model.
Note:

For template-based training, the maximum limit of Standard Forms document type is 157286400 bytes.

Azure AI Document Intelligence offers both pretrained and custom models to address a wide range of use cases for processing Standard Forms. This service is integrated with Document Automation and enables users to efficiently handle Standard Forms processing.

When you create custom extraction models, you must test the models by processing sample documents to assess the performance. Based on the test results, you might have to tweak or tune the model to achieve the desired results. After you achieve the desired results, you can use the model to process production documents.

Procedure

  1. In the Control Room, navigate to Manage > Learning Instances and click Create model.
  2. In the Projects window that opens, click Create project.
  3. Enter a name for the project, click Browse to upload documents to train the extraction model, and click Create.
  4. Define the field and table tags. See Create custom models in Document Automation using Standard Forms.
  5. Train the model. See Create custom models in Document Automation using Standard Forms.
  6. To test the extraction model, click the Analyze icon, click Browse for file to select a document, and click Run analysis.
    Note: You cannot edit an extraction model after associating it with a learning instance.