When you click the Document ID for a submission in the Documents table, you will be taken to the Document Output page. The Document Output page contains the original page image, the machine and manual transcriptions, and some additional details that may assist in debugging and performance investigation.
This page has three main sections: the document summary, the field list, and the image viewer.
Document summary
This section is at the top of the Document Output page and contains the aggregate stats of all pages in that document.
This section is hidden by default, so to view this section click on the blue downward arrow next to the document name.
Status
If there are outstanding Supervision tasks for the document, a link will appear that allows you to complete those tasks. Upon completing the task, you will be redirected back to the Document Output page.
If there are no available Supervision tasks, the status will be "Complete."
Field automation
The charts show the percentage of fields’ occurrences that were transcribed by the machine. Underneath each graph is the exact number of fields’ occurrences that the machine extracted. Note that field automation metrics only count non-table fields’ occurrences.
Field exceptions
Clicking the linked text in the Field Exceptions card will show a list of exceptions for the submission
Below is the list of possible field exceptions. A field may have more than one exception.
Exception | Cause |
|---|---|
Consensus Field Autotranscribed | This field was marked in the Layout Editor as requiring Consensus, but a Machine Only status was applied to the layout or submission. As a result, no Supervision Transcription task was created. |
Illegible Field | This field was marked as illegible in the Supervision Transcription task by a keyer. |
Field Character Limit Reached | The field’s transcription exceeded 2000 characters and was truncated to meet that limit. |
Normalization Error | The system had problems normalizing the field. |
Required Field Missing | This field was marked as required in the Layout Editor, but was not filled out in the form. |
Supervision Always Field Autotranscribed | The field was marked in the Layout Editor as requiring Supervision review, but a Machine Only status was applied to the layout or submission. As a result, no Supervision Transcription task was created. |
Supervision Required, But Disabled | The field required Supervision, but Supervision was disabled for the submission. |
Validation Override | A keyer overrode the formatting constraint of the field. |
Document output
Page viewer
To navigate between pages in the document, click on the page thumbnails to the left of the center page image.
Click on any of the blue boxes on the document to view the field output in the field list on the right panel.
Use the top toolbar to zoom in/out, see the original page image, and show/hide the field list.
Field list
The field list to the right of the page viewer displays all of the fields in the document and their extracted values.
To see the field on the image viewer, select the field name in the field list, and a line connecting to the field will appear.
All identified occurrences of a field appear in the field list.
To see table fields, select View Table to hide the field list and display the table output at the bottom of the page.
For nested tables, you have two View Table buttons - one for the parent table and another one for the child table.
If Manual Identification Supervision is disabled, and the machine has a low level of confidence in table cells’ predictions, the table cells will be stored in the table output with the Supervision Required, But Disabled exception. If Supervision is enabled, the table cells will undergo Supervision, and the identification of each cell will be attributed to either the machine or a user.
If a field was marked as a Duplicate during the layout-creation process, only the first instance of that field in the document will appear in the field list.
To view more information about a field, click the arrow next to the field name (>). Click See more to view the keyer who worked on the field, the data type, multiline/dropout settings, if the field is required for Transcription, and what Supervision tasks it is required to go through.
The default transcribed value is normalized. Switch between Raw/Normalized by clicking the dual arrows (⇌) at the top of the field list. Table values will show up regardless of the Raw/Normalized filter selection.
If a field has multiple bounding boxes, the system concatenates the bounding boxes’ transcriptions into a single value.
To filter the fields displayed, click the View All drop-down list to filter fields based on field exceptions, table fields, manual/machine transcriptions, manual/machine identifications.