...
...
The pending actions
This will show the user how many actions are still required on the document level and on the page level. The actions on the page level are further broken down in (1) actions per table (see below) and (2) unrecognized lines.
You can see in the screenshot above that there are 7 pending page actions; seeing as this is only a one page document, the number of page actions is the same as the number of document actions.
...
The data tables
There are two types of data tables
...
You can access the different tables through the tabs in the data view. Each tab will hold a badge showing the number of actions that the user is still expected to take on the table. Please refer to Exception-based review for an explanation on how to navigate quickly from action to action.
The Fields table: individual fields
In this table, we show the different fields of interest that have been identified, in addition to placeholders for mandatory data that has not been found yet.
...
In the Content column, you'll find the information extracted out of the document for that label.
Other tables: tabular information
For tabular data, only one column header is fixed upfront: the row type. This behaves very similarly to the label for the fields table.
...
The 'x' and 'v' now don't just reject or approve one field, but an entire row. If there's a row with information that was detected in the footer of the document for example, and is nog needed in the final output result, you simply click the 'x' and the row is deleted.
Bulk actions
At the bottom of the tables, you will find the "Validate all" and "Reject all" buttons (see screenshot at the indiviual fields explanation). These will validate or reject all individual lines in the document.
...
Sections
Sections visualise document splitting: when a document is made up of multiple logical sections, then we want to make sure we can reflect this both in what you see on the screen and in what we eventually send back as the final result.
...