The DATAIA Institute
Selected by the Agence Nationale de la Recherche (ANR) as part of the « Investissements d'Avenir» programs, the DATAIA Institute aims to bring together and structure multidisciplinary expertise of great scope and visibility to better meet the major challenges of data science, artificial intelligence and their applications through the decompartmentalisation between mathematics, computer science and legal, economic and social sciences.
At its launch, it mobilizes more than 130 researchers and teacher-researchers from fourteen institutions: University of Paris-Saclay, Inria, CEA, CentraleSupélec, CNRS, Ecole polytechnique, ENSAE, HEC, IFP-EN, IMT, INRA, Université Evry-Val d'Essonne, Université Paris-Sud, Université Versailles - Saint-Quentin-en-Yvelines; as well as an industrial affiliation program enabling close collaboration with companies.
Data are collected and processed in all areas of private and professional life.
In the fields of transport or tourism, the agile and intense use of data is already making it possible to challenge the leading companies of the 21st century. Research and innovation through data have enabled the development of an entirely new economy.
These intelligent technologies and services have destabilized many traditional industries and are disrupting our lifestyles, the functioning of our organizations and our entire social model. While data-based business models are radically changing the landscape, they face major scientific and technological challenges (capturing, storing, processing exponentially growing data) and raise critical social, legal and ethical issues.
Just as data is ubiquitous, algorithms for managing and analyzing this data are becoming increasingly present and inherent in intelligent digital services.
Due to this duality between data and algorithms, mastering the robustness and stability of algorithmic systems is a major challenge.
Data analysis evolves from describing the past to predictive and prescriptive analysis. Depending on the country, vigilance has been created around sensitive data (privacy) and/or algorithms (prescriptive context). Questions then arise about the neutrality, equity, non-discrimination, loyalty, security and bias of these algorithmic systems due to the informational asymmetry between the producers of these digital services and their consumers, whether they are citizens or professionals (B2C or B2B). Thus, the transparency, accountability and explainability of algorithmic systems become critical properties.
Digital transformation based on data science and artificial intelligence presents scientific, technological, legal, economic and ethical obstacles that are totally interdependent in the current context.