知识图谱和AI智能体协同驱动的机器学习课程助学助教模式探索

Exploration of a Teaching and Learning Assistance Model for Machine Learning Courses Collaboratively Driven by Knowledge Graphs and AI Agents

摘要: 针对当前机器学习课程中教学资源碎片化、教师难以实时掌握学生学习情况的问题，本文提出“以知识图谱为中枢、AI智能体为双向接口”的助教助学一体化新范式。通过超星平台的知识图谱功能构建知识点图谱并将其与课程目标进行关联；为每个知识点配置多样化的学习资源与练习题库以方便知识掌握程度的量化分析；在此基础上基于讯飞星辰Agent开发平台构建助学助教智能体。应用表明，该模式不仅有效提升了学生的学习效率与目标清晰度，更赋能教师实现基于数据的教学决策与动态调整，形成了教学相长、闭环优化的新型教学模式。

Abstract: In response to the current issues of fragmented teaching resources in machine learning courses and the difficulty for instructors to grasp students’ learning progress in real time, this study proposes a novel integrated paradigm of teaching and learning assistance characterized by “knowledge graph as the central hub and AI agent as a bidirectional interface”. Using the knowledge graph functionality of the Chaoxing Platform, a knowledge point graph was constructed and linked to course objectives. Diverse learning resources and exercise question banks were configured for each knowledge point to facilitate quantitative analysis of knowledge mastery. On this basis, a teaching and learning assistance agent was developed utilizing the iFlytek Agent Development Platform. Application results demonstrate that this model not only effectively enhances students’ learning efficiency and goal clarity but also empowers instructors to make data-driven teaching decisions and dynamic adjustments, ultimately forming a closed-loop, optimized teaching model that promotes mutual growth between teaching and learning.