Extracts information from forms and images into structured data based on a model created by a set of representative training forms.