An architecture for document management

The main goal of this work is to investigate a computational architecture for a document management environment. Its purpose is to digitalize and extract information from documents of any type, transforming them into structured electronic documents. The environment is divided into specification and extraction modules. In the first module, the user performs the document specification, capturing physical, semantic and process information from the document. The extraction module uses this information, in order to recognize documents of the same class as the specified one. This environment offers the possibility of visualizing the classification results and to correct eventual mistakes it has made. Also, it allows document reconstruction from physical and semantic information captured in the specification module.