To view/modify data in the Document Table, click the name of the Document table in the Dataset listing screen.

Add Data

You cannot directly add data to the Document Table.

Data is only added by the task with Doc Reader Node. When the task is executed, the fields are extracted from the document and populated in the columns of the Document Table.

Edit data

Click the icon next to the required row. The Document familiarization window opens. Based on the execution Status of the document processed, perform the following actions.

1. If Status is NEW_TYPE, you can familiarize the fields from the document in the left panel. SAVE AND APPROVE the extracted fields in the right panel.
2. The extracted fields are populated in the respective columns. The Status gets updated to EXTRACTED SUCCESSFULLY.
3. If Status is EXTRACTED_SUCCESSFULLY, you can view the document in the left panel and the extracted fields in the right panel.

4. If the validation fails in Post Processing Task, the status gets updated to MANUAL_INTERVENTION_VALIDATION_FAILED.

   • If the validation failed due to wrong field extraction, you can retrain the document from the left panel and update the Fields in the right panel. The changes are applied only for that particular document and are not saved at category level.
   • If the validation failed due to addition of new labels, you can set the labels for the fields to be extracted as Pseudonyms that are saved at the category level.
     

   

5. If the Status is MANUAL_INTERVENTION_FOR_REVIEW the familiarization window is read only. You cannot edit the fields in right panel.

Auto Processing

Click the icon and enable the Auto Processing option.
For Invoice and Bill of Lading schema the values for predefined columns get predicted by the ML engine. Using the predicted data, the new categories are auto approved without a need for manual intervention.
The status changes to EXTRACTED_SUCCESSFULLY automatically, if the validations are successful.
The data is extracted based on the machine learning algorithm, trained and delivered as part of the Product.
If any user-defined columns, they will not be extracted and will be blank.

Auto processing is applied only for the predefined schemas.

Other Options

To know more about the other options, click here.