Artificial Intelligence for the study if Charleville, Ré (Island), Tous and INED demographic data – DAI-CRETDHI

Historical demography and the social history of the family have long sought to measure phenomena, in time or space, by favoring methods of analysis based on aggregated corpora. Recent advances in artificial intelligence open up new perspectives for revisiting these themes, starting directly from the individuals who make up the populations studied. Observing on an individual scale and considering the maximum number of personal attributes available (sex, age, marital status, profession, etc.) and their singular combinations, makes it possible to create different aggregations (family, genealogies, professional, migration, etc.) . This project is based on the complementary skills of 4 laboratories and 2 companies, determined to implement an automatic structuring of entities (people and places) in time and space to process very large volumes of data. The challenges to be met are linked to 1) the mobilization of sources of various kinds, 2) the observation of different geographical levels (national - based on reasoned samples from INED surveys -, regional - Touraine and the island of Re-, and local with the case of the new town of Charleville) and 3) in the long term considered (16th-19th centuries). Ultimately, it will be a question of being able to offer, in an automated way, large-scale family reconstructions (intergenerational chaining) and the reconstruction of individual paths and solidarity networks. These productions are not the ultimate goal of the project, but constitute the essential points of support for the analyzes which will be carried out by the demographic historians of the team, whose ambition is to contribute to a better knowledge of the population of France from the 16th to the 19th century, thus filling a large gap in French historiography. It will also create an original and replicable methodological model for any territory with sources of a comparable nature.