AI助教在教学中的角色谱系研究——以《数学分析》课程为例

Research on the Role Spectrum of AI Teaching Assistants in Teaching—Taking the Course of Mathematical Analysis as an Example

摘要: 生成式人工智能推动高校混合式教学改革，《数学分析》课程存在学情诊断粗放、个性化辅导不足、AI 助教工具化碎片化应用等痛点，现有研究缺少针对专业课程的系统化角色设计。本文依托角色理论与支架式教学理论，厘清AI助教“功能”与“角色”的核心差异，遵循学科适配、全流程覆盖、知识图谱锚定、人机协同互补四项原则，构建“2个全程核心角色 + 5个分阶段专项角色”的角色谱系。以函数极限ε-δ定义教学为例展示多角色协同流程，搭建“数据层–角色层–知识层”三级框架阐释体系闭环运行逻辑。研究指出，AI助教教学效能取决于角色体系完整性、协同有序度与学科适配深度，而非单一技术功能强弱。本研究可为数学专业课程人机协同教学、智能助教体系搭建提供理论支撑与实践参照。

Abstract: Generative artificial intelligence advances the reform of blended teaching in universities. The course Mathematical Analysis is plagued by rough learning diagnosis, inadequate personalized tutoring and tool-oriented and fragmented application of AI teaching assistants, while existing studies lack systematic role design for specialized courses. Based on role theory and scaffolding instruction theory, this paper distinguishes AI teaching assistants’ “functions” and “roles”. Following four principles of discipline adaptability, full-process coverage, knowledge graph anchoring and human-AI coordination, it constructs a role spectrum of “two whole-process core roles + five stage-specific special roles”. Taking the ε-δ definition of function limits as a teaching case, this paper demonstrates the multi-role collaborative process and interprets the closed-loop operation logic via a three-level framework of “data layer-role layer-knowledge layer”. The research reveals that the effectiveness of AI teaching assistants relies on the completeness of the role system, the orderliness of collaboration and the depth of discipline adaptability, rather than the strength of a single technical function. It provides theoretical support and practical reference for human-AI collaborative teaching in mathematics major courses and the construction of an intelligent teaching assistant system.