Construction of a multimodal learning analysis model in the context of smart classrooms

Key words: 
• multimodal learning analysis /
• smart classroom /
• model design /
• personalized learning /
• teaching decision making

Abstract: In contemporary education,the prevalence of the test-based education model has led to a single,rigid learning assessment,while neglecting the personalized learning needs of learners. The rise of the smart classroom and multimodal learning analytics provides a solution to this problem. In the smart classroom,learners interact with instructors,peers,technology,and learning aids through multisensory channels,generating a large amount of multimodal data. The collection and analysis of this multimodal learning data can be further supported with the help of key technologies in the smart classroom. In order to understand this analysis process,the study constructs a multimodal learning analytics model grounded in the smart classroom,based on the elaboration of the concepts and characteristics of multimodal learning analysis and the smart classroom,and by drawing on the analysis of existing models. The model is centered on a six-step multimodal learning analytics cycle of establishing goals,collecting data,processing data,analyzing data,providing feedback,and implementing interventions,and also includes key elements such as stakeholders,the smart classroom technological environment,multimodal data types,and collection devices. The study explores the specific elements and practical operational requirements of each step of the model,with a view to providing theoretical support for the development of a learning analytics system in the smart classroom at a later stage,and then providing a more comprehensive and personalized assessment method for classroom teaching and promoting the intelligent and personalized development of education.