Detecting and quantifying uncertainty in the Studium Parisiense corpus

Master’s internship carried out at the LIP6 (Sorbonne Université & CNRS), February–August 2026, supervised by Camelia Constantin, Raphaël Fournier-S’niehotta and Cédric du Mouza.

Studium Parisiense gathers more than 20,000 records on the members of the schools and university of Paris from the 12th to the 16th century. Drawn from incomplete medieval sources, these records mix established facts, hypotheses and approximate values. The thesis evaluates language models on detecting and qualifying the uncertainty their editors expressed, and converts the records into a knowledge graph following the DAPHNÉ model.

Recommended citation: Maugin, N. (2026). Construction d'un graphe de connaissances et quantification de l'incertitude dans le corpus Studium Parisiense. Mémoire de master, Université Paris Cité.
Download Paper