Datafying the ‘webs of knowledge’ around the Old University of Leuven (STUDIUM.AI)

STUDIUM.AI will develop a research infrastructure to examine the ‘webs of knowledge’ around the Old University
in Leuven (1425-1797) from a data- and AI-driven perspective. As a first objective, STUDIUM.AI will allow 8
teams in 4 faculties to collaborate within the same sustainable LOD-environment on data of students, scholars, texts, ideas and images in premodern Leuven, a major knowledge hub at Europe's crossroads. By its interdisciplinary scope, the RI overcomes the methodological divide between our current texts-, personsand image-based datasets. Secondly, an HTR-model will be developed to generate full-text from handwritten manuscripts containing lecture notes of Leuven students. The resulting corpora will be mined and linked through NLPprocesses
of Named Entity Recognition (NER) and Named Entity Linking (NEL). Thirdly, as a result, STUDIUM.AI envisages to create a data pool of about 166,000 records, which is exponentially larger than the current 8 datasets stemming from projects or heritage initiatives (between 400 and 10,000 entries). These ‘big data’ will allow to integrate recent AI-driven algorithms in network analysis to detect ‘clusters’ which remain invisible within individual disciplines and projects, or ‘hidden’ because of power, gender or migration. Hence, STUDIUM.AI offers a methodological reflection on how DH- and AI-methods can help to rethink the transmission of knowledge in past and present within an intersectional and inclusive analysis.