Recent developments in systematic musicology, incorporating data from cognitive sciences, have significantly transformed the way we approach temporal musical works. In musicology, this shift has led to the emergence of cognitive musicology, and more specifically, "cognitive analysis", which has introduced major advancements in how musical forms and structures are represented. The computational implementation of these new concepts could revolutionize the handling of digital data flows by emphasizing the importance of second-order algorithmics, capable of modeling key aspects of cognitive strategies activated by musical structures. Such algorithmics would enable the proposal of new representations, particularly self-representations of its own behavior, the significance of which for musical works remains largely unexplored. All these procedures could be subjected to psycho-cognitive validations, notably through online systems involving volunteer listeners in participatory research.
The objective of this thesis is thus to propose a computational model of these new ideas from cognitive musicology and to conduct the necessary computational and musicological experiments to validate the concepts and their concrete implementation.