Training Documents

After completing the document definition process, you will be directed to the document editor dashboard. Here, you can proceed with training by clicking on the document that requires further processing.

The Document Training Screen will open, presenting you with various options. In the middle panel, you can view the document itself, while the right panel displays the fields captured by the machine learning system. The left panel allows you to classify and declassify the documents for better organization.

To address missing fields in the extracted data:

1. Click on the Missing field located at the bottom left corner.
2. Locate the missing field, and then click the Extract button associated with it.
3. Drag and select the label or data you want to extract for the missing field.
4. Confirm the extraction by clicking the Save and Update button.

In the case of incorrect fields in the extracted data:

1. Right-click on the incorrect label to prompt a menu.
2. select the Area Selector option and choose Text.
3. Drag and select the correct label or data for the field.
4. Identify the appropriate field type (for example, name, date, etc.) for the selected data.
5. Confirm the correction by clicking the Save and Update button.

Training of Tables

Prerequisite:

To store the extracted data from tables, it is necessary to create a separate Business Object containing attributes corresponding to the table fields and link it to the main Business Object.

1. Normally, data extraction from the table occurs automatically, but if it hasn't happened, right-click on the Table to be extracted.
2. Select the Area Selector option and choose Table.
3. Drag and select the correct data for the Table field.
4. Select the correct table for extraction from the displayed list of tables. When dealing with multiple tables, it becomes crucial to select the correct table from the displayed list for accurate data extraction.
5. If your table contains columns with descriptions, select each description to add a column. Repeat this process for all columns in the table.
6. Mark a text that represents the end of table. You can add text for start of table also.
7. Confirm the correction by clicking the Save and Update button.

Training of Checkboxes or Images

1. Right-click on the check-boxes to be extracted.
2. Select the Area Selector option and choose either Image or checkbox.
3. Drag and select the correct image or checkboxes.
4. Confirm the correction by clicking the Save and Update button.

All documents in formats like those already approved will be automatically extracted and approved.