• Funding : Artois, Région HdF
  • Start year :
  • 2021


The data mining field (or knowledge discovery from data) has seen significant developments over the last twenty years. This evolution has been driven particularly by important applications in various domains ranging from finance, to business, to the humanities and social sciences. It is a multidisciplinary field that includes artificial intelligence and at least the domain from which the data is derived. This thesis is an extension of the Mastodons QDoSSI project [1], involving computer sciences, humanities, and social sciences, whose main objective is the analysis of data dealing with international migration. These heterogeneous, multi-source data present various imperfections and pose a real challenge for data mining techniques.

The general objective of this thesis is to promote innovative tools and open-source software prototypes that address a double challenge:

(1) to fertilize new approaches based on a dialogue between Social Sciences and Computer Science for a better exploitation and understanding of migration and refugee data by sharing them with policy makers and civil society;

(2) to place the analyst or researcher in human and social sciences at the heart of the exploration with the will to co-construct the process that will lead him to discover new knowledge, to validate working hypotheses, and to better orient himself in the conduct of his field investigations.