AI classification
AI classification is used to automatically recognize and mark data that may require protection from further use of the document.
The system uses search and data recognition algorithms developed by BTC, as well as mechanisms based on artificial intelligence. The classifier analyzes the content and structure of the document, identifies selected types of information, and indicates the places where they were detected.
The recognized elements are marked directly in the document view and grouped by category. This allows the user to review the analysis result before proceeding to anonymization.
Classification scope
The system classifies, among others:
Personal data First name, last name, PESEL number, document number, face photo, fingerprints.
Organization data Tax identification number, REGON, KRS, position.
Contact details Phone number, email address.
Financial data IBAN, credit card, amount, amount written in words.
Signatures Handwritten signature, stamp, stamp with signature, writing.
Address data Street and number, postal code, city.
Identification numbers Land and mortgage register number, ISBN, ISSN, license plate, VIN number.
Other elements Date, URL, barcode, QR code.
Verification of results
After the classification is complete, the user should check all detected fields and make sure that the data requiring protection has been correctly marked.
If necessary, you can:
delete an incorrectly detected field,
change its position,
adjust the size of the selection,
save the changes made before proceeding to anonymization.
AI classification supports the user in finding data, but does not replace final verification.
Next stage
After reviewing and correcting the classification results, the user can proceed to the anonymization stage and decide which data categories should be permanently removed from the resulting version of the document.
More information can be found in the sections Document view and Anonymization.
Last updated