Learner Models for Personalized Intelligent Tutoring Systems

Nancy

Intelligent Tutoring Systems provide personalized and adaptive learning experiences, with the goal of complementing, or replacing, human instruction. Many existing approaches rely on heuristics such as the Leitner system, which are both effective in practice and easy to interpret and implement. However, such heuristics often lack the flexibility to adapt to diverse learning conditions, goals, or habits. In this talk, I will present a complementary approach based on learner modeling. These models aim to predict how learners acquire and retain knowledge, enabling the system to adapt teaching materials accordingly. I will illustrate this approach through two applications: vocabulary learning, where memorization is the central challenge, and the acquisition of simple grammatical rules, where assessing reasoning and generalization capabilities of the learner must be assessed.