Automatic Extraction of Business Logic from Digital Documents

Executable papers have been attracting attention of scientific publishers, as they may improve validation of their content by reviewers. The paper reports on the ongoing research on making regular PDF file submissions executable without a need to build sophisticated Web systems and allowing authors to use their customary authoring tools. A unified approach has been developed to enable extraction of useful information from the digital PDF document, either scanned printouts or born-digital content and illustrated with experiments involving BPMN diagrams.