Un écosystème adaptatif pour l’apprentissage des langues : Transformation intelligente et ajustement dynamique dans LinguaNeural Kids

Abstract

This thesis presents a scalable ecosystem for adaptive language learning, extendable across educational levels (primary to higher education, private and individual) and potentially to other subjects. Its core engine analyzes learner behavior, adjusts diffi-culty dynamically, rotates tasks, memorizes errors, and determines completion based on demonstrated mastery. The first implementation, LinguaNeural Kids, targets Algerian primary pupils (3rd–5th year). It structures the system across four coordinated appli-cations and embeds adaptive mechanics with early emotional intelligence (facial/vocal). The Apprentissage module automatically converts textual content into interactive ac-tivities, bridging authoring and adaptation while maintaining human validation. With a modular architecture, UML modeling, strong data-protection constraints, and an en-trepreneurial feasibility analysis, the project shows technical coherence and commercial promise, pending experimental, pedagogical, and ethical validation