Grand Tour digital. Digitization, Exploration, and Visualization of Early Modern Ego Documents from Educational Journeys Using Semi-Automated Editorial Methods

The research project presented in the poster experiments with methods for editing and visualising early modern travelogues using the handwritten text recognition (HTR) software Transkribus as well as Named Entity Recognition (NER). The focus is on the partially automated editing of mainly German-language travelogues from the early modern period (16th - 18th centuries) and the development of an HTR model specifically for travel literature. This model will be trained on a large number of authors' hands, the repeated use of foreign terms and quotations, as well as the high frequency of proper names (places, persons etc.) in the text corpus. After training this model, the project will investigate and test the reusability of the transcribed text and the associated annotations in relation to NER methods. Central questions in this context are the challenges and perspectives of exporting the data, as well as the possibilities of reusability with regard to information extraction and the creation of data visualisations. For further informations see the published texts (http://selbstzeugnisse.hab.de/edition_grand_tour) and the project blog (https://grandtourdig.hypotheses.org/).