The document view is the central interface in smartextract where you can review, validate, and refine extracted data from your documents.
This powerful workspace provides a comprehensive set of tools for ensuring the accuracy of your extracted information and customizing models to better suit your unique requirements.
The document view is divided into three primary sections:
left panel: Document thumbnail navigation displaying page previews for multi-page documents, allowing you to quickly jump between pages
center panel: Full document display with highlighted bounding boxes indicating recognized data fields
right panel: Structured display of all extracted fields organized by logical categories
smartextract’s AI models are highly accurate, but verification is sometimes necessary. The system makes it easy to identify and correct any extraction errors.
Visual verification: Bounding boxes in the document show exactly where each piece of data was extracted from
Field highlighting: Click on any extracted field in the right panel to see its corresponding location highlighted in the document
Navigation links: Click on field labels to jump directly to their location in the document
If you notice an incorrect extraction, you can easily fix it:
Click on the field value in the right panel
Edit the text directly in the field
The system automatically saves your correction
While smartextract provides several pre-built models, you can modify them to better align with your specific document formats:
Click the "Customize AI model" button in the top right corner
The customization interface will appear in the right panel
Field groups organize related information into logical sections:
Click "New field group"
Provide a descriptive name (e.g., "Certificate Information")
Click "Next" to begin adding fields
smartextract supports multiple field types to accommodate various data formats:
Click "Add field"
Enter the field name (e.g., "Register Number")
Select "Text" as the field type
Click "Confirm"
Text fields are ideal for extracting document numbers, names, addresses, and other textual information.
Click "Add field"
Enter the field name (e.g., "Ventilation Type")
Select "Multiple Choice" as the field type
Enter all possible options separated by commas (e.g., "Natural, Mechanical, Hybrid")
Click "Confirm"
Multiple choice fields are excellent for standardizing data where only certain values are valid, such as categories, statuses, or classifications.
Click "Add field"
Enter the field name (e.g., "Certification Date")
Select "Date" as the field type
Click "Confirm"
Date fields automatically format extracted dates in a consistent manner.
Click "Add Field"
Enter the field name (e.g., "Total Area")
Select "Quantity" as the field type
Click "Confirm"
Quantity fields ensure that extracted values are properly recognized as numbers for calculations and reporting.
You can create fields for information that isn't explicitly stated in the document but can be inferred:
Add a new field (e.g., "Signature Status")
Select "Multiple Choice" as the field type
Enter the possible values (e.g., "Signed, Unsigned")
Click "Confirm"
The AI model will analyze the document and make a determination based on visual cues and content.
If critical information isn't being extracted:
Click "Customize AI Model"
Add the missing field to the appropriate field group
Save the model and allow it to reprocess the document
Verify the field is now being correctly extracted
Review strategically: Focus on high-value fields and those with lower confidence scores first
Maintain consistent naming: Use clear, descriptive field names that match across similar documents
Group related fields: Organize fields logically to improve usability and extraction accuracy
Document model changes: Keep notes on customizations for team knowledge sharing
The Document View in smartextract offers powerful tools for reviewing, correcting, and customizing your data extraction process. By understanding how to effectively use these features, you can ensure the highest level of accuracy while tailoring the system to your specific document processing needs.
Whether you're using pre-built models or creating custom extraction templates, smartextract puts you in control of your data extraction workflow, enabling you to process documents with confidence and precision.