Analysis of Concentration Based on Multimodal Data in Synchronous Classrooms
Synchronous classrooms,the main approach to implementing the"Three Classrooms",have become an important measure to promote balanced development of education.Timely and accurate grasp of learners'concentration and taking effective measures for intervention are of great significance for enhancing the teaching quality of synchronous classrooms.However,in current synchronous classrooms,there are difficulties such as the inability to accurately feedback and adjust the concentration status of learners in different locations.There is an urgent need to provide precise concentration analysis and feedback schemes from a technical perspective.In order to achieve accurate feedback on the concentration status of learners in synchronous classrooms,the analysis of concentration in synchronous classrooms needs to sequentially solve the three core issues of"modeling","relationship"and"application",namely the calculation of concentration based on learners'data,the correlation among concentration features,and the intervention strategies for concentration due to different causes,so as to realize the regulation of concentration on learning.To this end,a dataset can be initially constructed based on the multimodal data such as visual and physiological data of individual learners in synchronous classrooms,from which the action and emotion features are extracted to form a concentration assessment model.Subsequently,the intrinsic dynamic evolution mechanism of group concentration is analyzed and an adaptive model for the evolution of group concentration is constructed.Finally,a"problem-intervention strategy"knowledge graph is established,and a concentration intervention model supported by human-machine intelligent collaboration and a precise teaching model for synchronous classrooms are erected,forming a complete chain of personalized learning services of"data perception—problem diagnosis—intervention strategy matching",which assists teachers in accurately identifying learners'problems and enhancing the effectiveness of classroom teaching.Further tests and optimizations will be needed in actual synchronous classrooms for the adaptability of the modeling indicators,analysis models,and intervention strategies,as well as the acceptance of teachers and learners.
NETL
NSTL
万方数据
