生成式AI赋能初中数学协同评价理论模型与实现路径研究

Research on the Theoretical Models and Implementation Pathways of Collaborative Assessment in Middle School Mathematics Empowered by Generative AI

摘要: 在教育数字化与人工智能深度融合的背景下，传统初中数学评价亟需向过程性、精准化与发展性转型。本研究基于生成式AI，参考人机协同评价理念，构建了一个面向初中数学课堂的生成式AI赋能的人机协同评价理论模型。该模型以数学核心素养为导向，设计“能力层–二级维度层–具体指标层”三级评价维度体系，并融合教师、学生与AI三方主体，形成“准备–评价–反思”三阶段闭环流程。通过结构化量表与大语言模型的协同，实现对数学抽象、逻辑推理、建模应用等高阶思维的精细化诊断与个性化反馈。结合“一次函数资费比较”案例，展示了模型在真实教学场景中的可操作性与应用潜力。研究表明，该框架有助于推动评价从“结果判定”转向“过程支持”，为构建“师–机–生”三元协同的智慧教育新生态提供理论支撑与实践路径。

Abstract: Against the backdrop of deep integration of educational digitalization and artificial intelligence, traditional evaluation in junior high school mathematics urgently needs to transform towards process-oriented, precise, and developmental paradigms. Based on generative AI and the concept of human-computer collaborative evaluation, this study constructs a theoretical model of generative AI-enabled human-computer collaborative evaluation for junior high school mathematics classrooms. Oriented by core mathematical competencies, the model designs a three-level evaluation system comprising “capability level-secondary dimension level-specific indicator level”, and integrates three stakeholders—teachers, students, and AI—to form a closed-loop process of “preparation-evaluation-reflection”. By combining structured rubrics with large language models, the framework enables fine-grained diagnosis and personalized feedback on high-order thinking skills such as mathematical abstraction, logical reasoning, and modeling application. A case study on “comparing charging plans using linear functions” demonstrates the model’s operability and application potential in real teaching scenarios. The research indicates that this framework helps shift evaluation from “outcome judgment” to “process support”, providing both theoretical foundations and practical pathways for building a smart education ecosystem of “teacher-machine-student” ternary collaboration.