Towards a Viability Approach for the Development of Task-Skill Mappings for the Validation of Prior Learning

Abstract

Inclusive learning aims to guarantee equitable access to educational opportunities by validating competencies acquired across diverse contexts. In Côte d’Ivoire, where a major share of the active population operates within the informal sector, the Validation of Prior Learning (Validation des Acquis de l’Expérience – VAE) constitutes a critical lever for certifying these skills. However, its operational implementation remains complex due to cumbersome evaluation procedures and the rigidity of certification frameworks. This paper proposes an innovative dynamic systems approach rooted in viability theory to automate and optimize the design of the “task-skill” mappings essential to VAE. By leveraging graph Laplacian operators and differential inclusions, this method models the propagation of knowledge flows within a bounded viability kernel. Experimental validation conducted on a cohort of 300 artisans from the Lagunes region demonstrates the robustness of the model: the stabilized profiles achieve high topological congruence with the official CQP (Professional Qualification Certificate) framework, yielding a Jaccard similarity index of 0.92 and a Cohen’s Kappa coefficient of 0.84, while ensuring rapid asymptotic convergence within 150 iterations. The results confirm that this viability approach improves the reliability of didactic diagnostics and reduces the temporal overhead of assessment sessions, thereby facilitating certification access for individuals from informal backgrounds and mitigating educational inequalities. Moving forward, this adaptive model paves the way for hyperparameter self-calibration and sectorial transposition to other evolving professional fields.