Linking University, City and Diversity
This interdisciplinary project applies data science techniques to qualitatively and quantitatively visualize the interaction between the University and the city of Leiden (Town&Gown) from 1575 onwards. One of the possible angles for follow-up research is diversity, especially from the international relationships built up over the centuries and the impact of the arrival of students and professors from outside the Netherlands on the city.
- Ariadne Schmidt
- Leiden University
Historisch Erfgoed Leiden, Historisch Leiden in Kaart, Universiteit Utrecht
Unlocking digitized sources enables new historical research
Collaboration with data scientists and heritage institutions makes it possible to link different types of sources, such as the register of professors, student entries in the Album Studiosorum, census records and map material (cadastral data) to persons and parcels of land, allowing the data to be queried and searched in conjunction.
Challenges from a data science perspective
Making digitized historical sources suitable is an area where machine learning techniques have been applied for several decades but where many challenges remain. In this context, the following questions are important:
- From raw data (images historical newspapers) to structured data in data model: article segmentation, recognizing entities in OCR text
- Network analysis on structured data (subgroups, interaction, centers of gravity)
- Diachronic changes in language use, perception (sentiment), and presentation (text mining)
- Visualization for retrieval / analysis
In the pilot project we initially focus on questions 3 and 4.
In the project, we focus on creating a basic infrastructure that will enable new historical research. That infrastructure will make it possible, through an interactive timeline and a map, to search, present and support analysis of structured data on persons, plots and events.
On Feb. 3, the project plan was presented to outgoing rector/chair Carel Stolker (who is supporting the project with a start-up grant) and new college president Annetje Ottow (see also: https://www.universiteitleiden.nl/en/news/2021/02/sculptures-provide-more-diverse-view-of-universitys-history)
With the start-up grant we will work on a visualization in time and space of available datasets on students and professors, and do a text mining analysis of historical newspapers from the KB. The intention is to present the project results during Leiden City of Science 2022.
The project is realized by Bachelor and Master projects by LIACS students in close collaboration with the Institute of History. In 2022, these were Rick Schreuder, Liam van Dreumel and Michael de Koning. A project team consisting of Richard van Dijk, Ben van Yperen and Michael de Koning is consolidating this base into a website with visualizations of professors and students enrolled at Leiden University. The visualizations will be available soon.