Rich wanted among other experiments to try and use MBL techniques on images and related data.


    1. It rapidly became clear that there were not enough images combined with descriptive data at the RACM to really get started with image retrieval in the way as originally intended.

    2. There exists a large quantity of papers, texts, reports and other documents that do contain such images and data, but there are no tools to analyse such documents to the level needed for experiments as envisaged above.

    3. So we decided to create a system that could collect data from texts and combine them with images.

    Open Boek