Shortfacts - February 2017

Learning from documents

Text recognition in automated document processing can now be optimized even more easily with Tangro. It is now possible to learn document content directly in the all-in-one mask - without changing the user interface. A bidirectional connection has been created between PDF and the SAP GUI of the Tangro mask.

This means that the assignment of document contents and field values can now be made visually visible and can therefore be recognized at a glance. New document content can be learned with a simple mouse click. In this way, the results of OCR processing can be improved even faster and more conveniently.

At the heart of the new learning functionality is a PDF viewer familiar from modern web browsers, which Tangro has developed specifically to improve text recognition. The PDF viewer is part of the all-in-one mask, in which the document image and the Tangro mask with all header and position data are displayed together. Users can now also learn documents in this familiar working environment.

This means that content in the PDF can be assigned to the field values in the Tangro mask and thus to the header, item and totals data in SAP via menu selection.

Conversely, field values can be highlighted in the PDF from SAP at the click of a mouse, making it easy to find position data, for example. The functions are so intuitive to use that no IT expertise or programming knowledge is required to continuously expand the pool of learning data.