A Framework for MOOC Learning Assessment based on Facial Expression driven by Knowledge Contextual Awareness

With the advancement of interactive hardware technology, intelligent cameras can be utilized to capture users' facial expressions for real-time assessment of Massive Open Online Course (MOOC) learning. However, existing learning assessment methods fail to consider the correlation between knowledge points and learning state, resulting in limited reliability when using facial expressions to evaluate online learning. Therefore, we propose a framework for MOOC learning assessment based on knowledge contextual awareness driven by facial expression analysis. In our manuscript, an offline pre-trained model is employed to automatically identify the knowledge points within MOOCs. Additionally, we design a dual-channel model that combines fine-tuning vision Transformer and SE-CNN for accurate recognition of facial expressions. Subsequently, the learning assessment is derived through facial expressions driven by knowledge contextual awareness. Finally, experimental results demonstrate the effectiveness of our proposed online learning assessment framework.