知识图谱在大学计算机基础课程教学中的应用

The Application of Knowledge Graphs in Teaching University Computer Fundamentals Courses

摘要: 针对大学计算机基础课程知识点复杂、学生认知差异较大、教学资源匹配效率低等问题，本文探讨了知识图谱赋能课程教改的理论基础、技术前沿、特点现状与应用路径。首先，剖析了知识图谱的语义表示、图结构生成、灵活性和可扩展性、可解释性等多个特点；然后，梳理了知识图谱增强检索(GraphRAG)、多模态知识图谱、知识图谱与大语言模型融合三大发展趋势；对比了传统人工多阶段构建流程与基于大语言模型的构建新范式，分析人工提取和大模型提取在抽取方法、模式构建、人工成本等方面的区别。重点阐述了知识图谱在课程教学中的四大应用：以图结构网络实现知识体系框架构建；通过语义检索实现知识点的快速查询与智能推荐；基于图拓扑生成个性化学习路径；依托知识点–资源–关系三层结构实现教学资源的智能匹配。最后，展望了大模型协同构建、多模态知识融合、跨课程图谱及基于知识图谱的智能答疑等未来发展方向。

Abstract: Addressing the challenges of complex knowledge points, significant differences in student cognition, and low efficiency in matching teaching resources in university computer fundamentals courses, this paper explores the theoretical foundation, technological frontiers, current characteristics, and application paths of knowledge graph-enabled curriculum reform. First, it analyzes several characteristics of knowledge graphs, including semantic representation, graph structure generation, flexibility and scalability, and interpretability. Then, it outlines three major development trends: knowledge graph-enhanced retrieval (GraphRAG), multimodal knowledge graphs, and the integration of knowledge graphs with large language models. It compares the traditional multi-stage manual construction process with the new paradigm based on large language models, analyzing the differences between manual extraction and large model extraction in terms of extraction methods, pattern construction, and labor costs. The paper focuses on four key applications of knowledge graphs in curriculum teaching: constructing knowledge system frameworks using graph-structured networks; enabling rapid querying and intelligent recommendation of knowledge points through semantic retrieval; generating personalized learning paths based on graph topology; and achieving intelligent matching of teaching resources based on a three-layer structure of knowledge points, resources, and relationships. Finally, it looks forward to future development directions such as collaborative construction of large models, multimodal knowledge fusion, cross-course graphs, and intelligent Q&A based on knowledge graphs.