The description is submitted to:


Session R4830 ("Can You Read This One?" New Advances and Issues in Digital Recognition of Printed/Handwritten Arabic Characters), 2017
As research group, we are working on an Ottoman economic history project with strong components of digital humanities. One of our crucial tasks is to digitise mid-nineteenth century Ottoman population and tax registers, which are kept in rik’a script. We are working on extracting demographic and economic data from these hand written sources and creating a machine readable dataset. Conventional optical character recognition technologies are not useful for our purposes. Therefore, we are hoping to develop bespoke intelligent character recognition techniques for semi-supervised data entry. If we could achieve this goal, and could apply newest and advanced techniques, such as, word spotting and computer vision, for Ottoman hand written documents, we could reach an unprecedented pace and precision of data entry and expand our dataset considerably.

Until now we have been suffering from the lack of interdisciplinary co-operation with data and computational scientist. As an Ottoman economic historian, it is an unsurmountable challenge for me to follow the ever expanding literature on digital character and object recognition methodologies, developed by computational scientists. To try to apply those methodologies in our field is simply beyond my digital and computational capabilities.

I would very like to get into a dialogue with experts in the field of computational sciences, to have a better understanding of possibilities and constraints of advanced digital humanities methods, relevant for Ottoman studies. To my understanding such an initial dialogue is a pre-requisite to initiate interdisciplinary research teams and design digital Ottoman history projects.