More about Free-Form Recognition

How does it work?

Document classification and data extraction is done using ‘rule sets’, making free-form recognition widely applicable. Rule sets are created and edited using the Form Designer, a GUI driven application included as part of the Form Recognition Module. These rule sets contain the information and relationships used to identify document types and data to be extracted. For example, an alphanumeric numeric sequence is found located in a position to the right of the term ‘invoice’ on a document. This document can then be classified as an ‘invoice’ and the invoice number extracted. Rule sets are saved as XML and are assigned to particular jobs, or batches of documents, within the PDF Compressor. Classification and data extraction then requires no manual intervention. Furthermore, the form designs themselves have statistical requirements for character recognition. If a value is not recognized or is considered suspect, it is forwarded automatically into the Form Validator where a human operator can key in or correct the output. This quality assurance enables up to 100 percent accuracy with appropriate corrections and eliminates manual data entry.

Additionally, LuraTech assist in designing your first rule sets, further reducing implementation effort and therefore leading to even quicker ROI.