This study focuses on the Facial Expression Recognition and Sentiment Analysis System V1.0 for learning situation analysis. It unfolds around the research of technical feasibility demonstration, expression classification extraction, and practical application verification. By improving the deep neural network model to enhance the accuracy of expression recognition, collecting procedural expression data to support learning situation analysis, and ultimately optimizing the system through teacher recognition surveys and student experience feedback, the study achieves the integration of technical research and teaching application.