Nurturing Self-aware Learning through Facial Expression Interpretation

A growing focus is on optimizing learning outcomes by supporting metacognitive monitoring, enabling learners to self-assess and adjust their learning strategies. Our goal is to support children in enhancing their metacognitive monitoring skills-the ability to precisely assess and regulate their learning processes. Research shows that targeted strategies can help children improve their skills, provided their imprecise metacognitive monitoring performance (MMP) is identified. However, conventional approaches for identifying imprecise MMP often have low performance and fail to provide the precision needed for effective interventions. To address this, we aim to leverage deep learning to analyze facial expressions as a means of estimating MMP. By interpreting subtle facial expressions, our proposed deep learning approach offers a promising alternative to conventional approaches, enabling more precise and timely estimations of MMP. Our provisional findings suggest significant implications for tailoring strategies for learners’ metacognitive monitoring.