Bridging the media gap from the Guthenberg´s world to electronic document management systems

The sender of a document usually decides who should receive his or her message. However because of the flood of information, a recipient should also be capable of deciding what kind of information he likes to receive. The author presents a prototype version of the system OfficeMAID which analyzes incoming printed documents in the purchasing department of a real company, extracts the relevant information, and automatically identifies the respective workflow. The paper discusses the existing weaknesses of document management systems, the various processing tasks of document analysis and understanding and shows experimental but promising results