Rapporten project volautomatische archiefontsluiting
Het eindrapport en de deelrapporten van het project Volautomatische Archiefontsluiting (Full Automatic Archival Access) zijn gepubliceerd.
"Aim of this project was to find out more about the possibilities of currently available proven technology to create useable machine-readable text out of analogue archival typed or hybrid documents. The project partners – Netwerk Oorlogsbronnen (Network War Collections), Nationaal Archief (National Archives of the Netherlands), Impact Centre of Competence and Centre of Language and Speech Technology – compiled a small non-representative sample set from the Centraal Archief Bijzondere Rechtspleging (National Archives). The tests focused on the best settings for scanning, pre-processing for Optical Character Recognition (OCR), transcription (OCR) and post-correction (in particular named entity recognition), using state-of-the-art software and common practice tools."