Résumé
Partial orders and lattices play a major role in the scientific method, especially for ordinal and nominal data classification and comparison. Lattice theory has been developed for many years, with seminal works by R. Wille, B. Ganter, M. Barbut, B. Monjardet, and many other researchers. Strong connections with philosophy, logic, linguistics, and data science have been investigated over the years. 2022 is the 40th anniversary of the foundational paper of R. Wille “Restructuring Lattice Theory: An Approach based on Hierarchies of Concepts”, which paved the way for the dissemination of Formal Concept Analysis (FCA). FCA galaxy, including Galois connections, concept lattices, and implication bases, provides a versatile and very powerful toolbox for addressing complex knowledge extraction and exploration. It has been expanded into a large range of approaches for dealing with numerical and interval data, temporal and spatial data, sequences, graphs, or multi-relational data. Patterns structures, Fuzzy FCA, Graph-FCA, or Relational Concept Analysis are examples of such extensions. In this talk, I survey FCA from its origins, recalling key milestones. I report typical and successful applications in social sciences, life science, ontology engineering, and software engineering with several focus. I also bring to the light the existing and potential combinations with other data science approaches in the search for efficient Knowledge Discovery approaches and services.