| 1,344 | 0 | 673 |
| 下载次数 | 被引频次 | 阅读次数 |
人工智能重构教育系统背景下,构建中小学人工智能课程知识图谱是智能化人才培养的重要举措。然而,现有研究多集中于高等教育领域,缺乏与核心素养目标的深度融合,难以满足中小学人工智能教育需求。为此,该文以人工智能素养框架为指导,依托广州市中小学人工智能课程教材,采用自顶向下方法构建面向中小学的课程知识图谱。为验证其有效性,研发课程知识图谱增强的大模型问答系统,并通过人工评估测试系统性能。研究结果表明,课程知识图谱通过结构化知识注入机制,显著提升了大语言模型在人工智能素养的情感、思维、知识三个维度上的问答表现。该文通过课程知识图谱与大语言模型的融合应用,探索其在教育场景中的增益效应,实现从知识体系重构到工程实践的范式跃迁,为人工智能素养教育的规模化推广提供了理论与实践耦合的技术框架。
Abstract：As artificial intelligence(AI) reshapes modern education, the construction of AI curriculum knowledge graphs(CKGs) for K-12 schools has emerged as a pivotal strategy for cultivating AI-ready competencies. Current research, however, predominantly focus on higher education, often lacking integration with established competency frameworks and thus failing to address the specific pedagogical needs of K-12 AI education. To bridge this gap, this study constructed a dedicated CKG for K-12 learners. Its design was grounded in a formal AI literacy framework and implemented via a top-down approach, utilizing curriculum materials from Guangzhou's K-12 AI program. This study further developed a CKGenhanced, LLM-based question-answering(QA) system and evaluated its efficacy through rigorous manual testing. The results demonstrate that the CKG, through structured knowledge injection mechanism, significantly enhances the performance of LLM-based QA systems across the affective, cognitive, and knowledge dimensions of AI literacy. This work explores the synergistic effects of integrating CKGs with LLMs in educational contexts. It enables a paradigm shift from knowledge-system reconstruction to engineering practice, offering a dual-loop framework that interweaves theoretical rigor with pragmatic implementation to advance AI literacy education.
[1]中华人民共和国教育部.教育部部署加强中小学人工智能教育[EB/OL].http://www.moe.gov.cn/jyb_xwfb/gzdt_gzdt/s5987/202412/t20241202_1165500.html,2024-12-02.
[2]中国教育学会.中小学人工智能通识教育指南(2025年版)[EB/OL].https://www.cse.edu.cn/index/detail.html?category=31&id=4240,2025-05-12.
[3]中国教育学会.中小学生成式人工智能使用指南(2025年版)[EB/OL].https://www.cse.edu.cn/index/detail.html?category=31&id=4242,2025-05-13.
[4]祝智庭,戴岭等.新质人才培养:数智时代教育的新使命[J].电化教育研究,2024,45(1):52-60.
[5]王同聚.中小学人工智能课程教育实践策略的设计与实施[J].现代教育技术,2024,34(12):95-104.
[6]王全亮,黄剑华等.我国人工智能教育研究现状、热点与趋势——基于Cite Space和VOSviewer的可视化分析[J].教育与装备研究,2021,37(10):6-13.
[7]中华人民共和国教育部.教育部发布4项行动助推人工智能赋能教育[EB/OL].http://www.moe.gov.cn/jyb_xwfb/xw_zt/moe_357/2024/2024_zt05/mtbd/202403/t20240329_1123025.html,2024-03-28.
[8]张汉玉.基于共词分析的中小学人工智能教育研究热点及发展态势[J].教育与装备研究,2023,39(4):9-14.
[9]刘凤娟,赵蔚等.基于知识图谱的个性化学习模型与支持机制研究[J].中国电化教育,2022,(5):75-81+90.
[10]吴杨,吕钰琪等.知识图谱驱动智能学习的内生逻辑[J].中国电化教育,2025,(2):122-130.
[11][28]董晓晓,周东岱等.学科核心素养发展导向下教育领域知识图谱模式构建方法研究[J].电化教育研究,2022,43(5):76-83.
[12]Ji S,Pan S,et al.A survey on knowledge graphs:Representation,acquisition,and applications[J].IEEE Transactions on Neural Networks and Learning Systems,2021,33(2):494-514.
[13]李惠乾,钟柏昌.教育知识图谱:研究进展与未来发展——基于2013—2023年中文核心期刊载文的分析[J].计算机工程,2024,50(7):1-12.
[14]刘梅娟,余冰冰等.基于知识图谱的智能会计课程体系构建研究[J].财会通讯,2024,(13):160-164+171.
[15]郭宏伟.基于智能教育的高校在线课程知识图谱构建研究——以中国医学史为例[J].中国电化教育,2021,(2):123-130.
[16]丁子恒,丁子轩等.知识图谱在“物化生”概念教育中的研究与应用——以高中教材为例[J].教育与装备研究,2022,38(8):60-64.
[17]谢幼如,陆怡等.知识图谱赋能高校课程“教-学-评”一体化的探究[J].中国电化教育,2024,(12):1-7.
[18]杨丽萍,辛涛.核心素养十年探索:回顾反思与未来展望——专访北京师范大学中国基础教育质量监测协同创新中心辛涛[J].教育学报,2025,21(3):3-18.
[19]文森,钱力等.基于大语言模型的问答技术研究进展综述[J].数据分析与知识发现,2024,8(6):16-29.
[20]Wang B,Yue X,et al.Can Chat GPT defend its belief in truth?Evaluating LLM reasoning via debate[DB/OL].https://arxiv.org/abs/2305.13160,2025-08-01.
[21]Ji Z,Lee N,et al.Survey of hallucination in natural language generation[J].ACM Computing Surveys,2023,55(12):1-38.
[22]Baek J,Aji A F,et al.Knowledge-augmented language model prompting for zero-shot knowledge graph question answering[DB/OL].https://arxiv.org/abs/2306.04136,2025-08-01.
[23][43]王永,秦嘉俊等.融合知识图谱和大模型的高校科研管理问答系统设计[J].计算机科学与探索,2025,19(1):107-117.
[24]Guo T,Yang Q,et al.Knowledgenavigator:Leveraging large language models for enhanced reasoning over knowledge graph[J].Complex&Intelligent Systems,2024,10(5):7063-7076.
[25]马杰,孙望淳等.基于相关性提示的知识图谱问答[J].软件学报,2025,36(9):4056-4071.
[26][42]孙岩,周立新等.知识图谱和大语言模型在航天测控问答系统中的融合应用研究[J].上海航天(中英文),2024,41(5):178-184.
[27]陈静,曹智勋.对抗幻觉:垂直领域中大语言模型的应用策略探讨——以中医知识问答领域为例[J].数据分析与知识发现,2025,9(4):1-13.
[29]Ng D T K,Leung J K L,et al.AI literacy:Definition,teaching,evaluation and ethical issues[J].Proceedings of the Association for Information Science and Technology,2021,58(1):504-509.
[30][32] Lintner T.A systematic review of AI literacy scales[J].npj Science of Learning,2024,9(1):50.
[31]钟柏昌,刘晓凡等.何谓人工智能素养:本质、构成与评价体系[J].华东师范大学学报(教育科学版),2024,42(1):71-84.
[33]Noy N F,Mc Guinness D L.Ontology development 101:A guide to creating your first ontology[EB/OL].https://www.cs.upc.edu/~jvazquez/teaching/sma-upc/docs/ontology101.pdf,2025-04-29.
[34]T/CSE 001—2021,中小学人工智能课程开发标准[S].
[35]中小学人工智能课程指南课题组,江波.中小学人工智能课程指南[J].华东师范大学学报(教育科学版),2023,41(3):121-134.
[36]赵呈领,黄志芳等.基于初中物理课程的学科领域本体库构建研究[J].电化教育研究,2014,35(8):64-70+94.
[37]田卫民,武泽尧等.知识图谱赋能专业课程深度开发的创新实践[J].中国大学教学,2025,(5):16-25+63.
[38]李晓理,刘春芳等.知识图谱与大语言模型协同共生模式及其教育应用综述[J].计算机工程与应用,2025,61(15):1-13.
[39]Liang X,Wang H,et al.Controllable text generation for large language models:A survey[DB/OL].https://arxiv.org/abs/2408.12599,2025-08-01.
[40]Jiang J,Zhou K,et al.StructGPT:A general framework for large language model to reason over structured data[DB/OL].https://arxiv.org/abs/2305.09645,2025-08-01.
[41]Wu Y,Hu N,Bi S,et al.Retrieve-rewrite-answer:A KG-to-text enhanced LLMs framework for knowledge graph question answering[DB/OL].https://arxiv.org/abs/2309.11206,2025-08-01.
[44]Guo Q,Cao S,et al.A medical question answering system using large language models and knowledge graphs[J].International Journal of Intelligent Systems,2022,37(11):8548-8564.
[45]Chang Y,Wang X,et al.A survey on evaluation of large language models[J].ACM Transactions on Intelligent Systems and Technology,2024,15(3):1-45.
基本信息:
中图分类号:G633.67;G434
引用信息:
[1]黄景修,郑孜譞,赖飞宇,等.素养导向的中小学人工智能课程知识图谱构建与应用研究[J].中国电化教育,2026,No.469(02):46-52+59.
基金信息:
2022年国家自然科学青年基金项目“面向机器理解数学应用题的命题抽取方法研究”(项目编号:62207015); 2022年度教育部人文社会科学研究青年基金项目“人工智能支持下的在线讨论行为分析及适应性干预方法研究”(项目编号:22YJC880021)阶段性研究成果
2026-02-10
2026-02-10