Contextual and assisted interpretation of digitized fonds : application to sales registers from the 18th century

Fonds, also called historical document collections, are important amounts of digitized documents which are difficult to interpret automatically: usual approaches require a lot of work during design, but do not manage to avoid producing many errors which have to be corrected after processing.To cope with those limitations, our work aimed at improving the interpretation process by making use of information extracted from the fond, or provided by human operators, while keeping a page by page processing.We proposed a simple extension of page description language which permits to automatically generate information exchange between the interpretation process and its environment. A global iterative mechanism progressively brings contextual information to the later process, and improves interpretation.Experiments and application of those new tools for the processing of documents from the 18th century showed that our propositions were easy to integrate in an existing system, that its design is still simple, and that required manual corrections were reduced.