面向教学改进的数学课堂评价智能体:框架搭建与效果评价OA

Classroom teaching evaluation plays a pivotal role in driving the transformation of classroom paradigms,fostering teachers'professional development and elevating teaching quality.However,current classroom teaching evaluation generally relies on experienced teachers,making it difficult to meet the needs of regular teaching improvement.Based on analyzing 149 expert commentaries on teaching cases,we have refined a four-dimensional evaluation framework covering Content Knowledge,Student Learning,Teaching Strategies and Learning Outcomes.Drawing on the mathematics curriculum standards,textbooks and high-quality evaluation corpora as the knowledge base,we have developed a multi-agent-based automated evaluation system for classroom teaching.Benefiting from the construction of an external knowledge base and the multi-agent collaboration mechanism,the intelligent agents can more accurately identify problems in classroom teaching and put forward in-depth,actionable improvement suggestions compared with DeepSeek,a general large language model.These suggestions can serve as a reference for front-line teachers in their teaching improvement and expand the knowledge boundaries of school teaching and research activities.This system is expected to provide low-cost and highly accessible support for classroom teaching evaluation,thereby contributing to the improvement of classroom teaching quality and the balanced development of education.