Table Training window
Use this window to define table labels and table columns for table extraction.
All documents in Table Extraction Set for the selected class are displayed in this window. When the window is first opened, all documents are expanded and the currently selected document is displayed in bold in the document list. Use the expand and collapse icons to view or hide the detected tables in each of the training documents. In the same header row as the expand and collapse buttons there is also an icon that indicates if a document is included in the table training. Either the Include for Training icon is displayed or the Exclude from Training icon is displayed. Click on this icon to change a document training status as needed. When training is performed within this window, assigned labels and predicted labels are still provided for any excluded documents. However, this information is not used for table training. It is simply displayed as additional information.
This window has a menu and toolbar displayed at the top of the window so that you can access common functions.
For each document the window is separated into two distinct halves.
Left pane |
Right pane |
---|---|
The left pane displays all documents with detected tables. Whatever is selected in the left pane is displayed in the image view on the right pane. When you select a specific document the following information is available for the document and its detected tables:
|
The right pane displays the selected document page in an image viewer along with the following behavior:
|
When a table is selected, the following information about the columns in that table is displayed:
|
The following toolbar provides the settings for editing tables on a document and then adding it to your Table Extraction Set. These same settings are available in the menu.
Setting |
Name and Description |
---|---|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
Related topics: