| 816 | 0 | 755 |
| 下载次数 | 被引频次 | 阅读次数 |
随着人工智能技术与教育的深度融合进程加快,教育人工智能模型在应用过程中面临的信任与风险挑战日益凸显。针对这一关键问题,该研究创新性提出了“基础层-解释层-服务层”三层架构的可信教育人工智能模型框架。基础层重点解决教育数据的安全共享机制与算法公平性保障问题;解释层着力提升模型兼具教育学意义与技术实现的双重可解释性;服务层则致力于增强模型面向教育用户的透明度与可问责性。为验证该框架的实践价值,该研究选取教育领域的学习者模型作为典型案例,系统阐述了可信学习者模型的构建方法论与实现路径。最后,该研究进一步展望了可信教育人工智能领域未来需要重点突破的关键方向与发展路径。
Abstract：With the acceleration of the deep integration of artificial intelligence and education, the trust and risk challenges faced by the educational artificial intelligence model in the application process are increasingly prominent. To solve this key problem, this study innovatively proposed a three-tier architecture of “foundation layer-interpretation layer-service layer” framework for trustworthy artificial intelligence model in education. The foundation layer focuses on the security sharing mechanism of education data and the fairness guarantee of algorithm; The interpretation layer aims to improve the interpretability of the model, which has both pedagogical significance and technological realization; The service layer is committed to enhancing the transparency and accountability of the model for educational users. To verify the practical value of the framework, this study selected the learner model in the field of education as a typical case, and systematically elaborated the construction methodology and implementation path of the trustworthy learner model. Finally, this study further looks forward to the key directions and development paths that need to be broken through in the field of trustworthy artificial intelligence in education.
[1]卢宇,余京蕾等.基于大模型的教学智能体构建与应用研究[J].中国电化教育,2024,(7):99-108.
[2]卢宇,章志等.可解释人工智能在教育中的应用模式研究[J].中国电化教育,2022,(8):9-15+23.
[3]教办[2025]3号文件,教育部等九部门关于加快推进教育数字化的意见[Z].
[4]Li B,Qi P,et al.Trustworthy AI:From principles to practices[J].ACM Computing Surveys,2023,55(9):1-46.
[5]Independent High-Level Expert Group on Artificial Intelligence.Ethics guidelines for trustworthy AI[EB/OL].https://digitalstrategy.ec.europa.eu/en/library/ethicsguidelines-trustworthy-ai,2019-04-08.
[6]ISO/IEC TR 24028:2020,Information technology-Artificial intelligence-Overview of trustworthiness in artificial intelligence[S].
[7]OECD.OECD Principles on AI[EB/OL].https://www.oecd.org/going-digital/ai/principles,2019.
[8]中华人民共和国科学技术部.《新一代人工智能伦理规范》发布[EB/OL].https://www.most.gov.cn/kjbgz/202109/t20210926_177063.html?ref=salesforceresearch.63.html,2021-09-26.
[9]Vincent-Lancrin S,Vlies R V D. Trustworthy Artificial Intelligence(AI)in Education:Promises and challenges[R].OECD Education Working Papers,2020.
[10]Nguyen A, Ngo H N, et al. Ethical Principles for Artificial Intelligence in Education[J].Education and Information Technologies,2023,28(4):4221-4241.
[11]Hong Y,Nguyen A,et al.Data Ethics Framework for Artificial Intelligence in Education(AIED)[A].2022 International Conference on Advanced Learning Technologies(ICALT)[C].Bucharest:IEEE,2022.297-301.
[12]江波,丁莹雯等.教育数字化转型的核心技术引擎:可信教育人工智能[J].华东师范大学学报(教育科学版),2023,41(3):52-61.
[13]武法提,李坦等.可信学习分析构建:风险桎梏、理论纾解与实践解蔽[J].远程教育杂志,2024,42(5):12-22.
[14]杨俊锋,褚娟.人工智能教育应用的伦理风险和规范原则[J].中国教育学刊,2024,(11):21-27.
[15]王萍,田小勇等.可解释教育人工智能研究:系统框架、应用价值与案例分析[J].远程教育杂志,2021,39(6):20-29.
[16]Darvishi A,Khosravi H,et al.Incorporating AI and Learning Analytics to Build Trustworthy Peer Assessment Systems[J].British Journal of Educational Technology,2022,53(4):844-875.
[17]Jackson S,Panteli N.Trust or Mistrust in Algorithmic Grading?An Embedded Agency Perspective[J].International Journal of Information Management,2023,69:102555.
[18]Delgado Kloos C,et al.H2O Learn-Hybrid and Human-Oriented Learning:Trustworthy and Human-Centered Learning Analytics(Ta HCLA)for Hybrid Education[A].2022IEEE Global Engineering Education Conference(EDUCON)[C].Tunis:IEEE,2022.94-101.
[19]卢宇,章志等.可解释自动批阅模型构建与应用[J].开放教育研究,2023,29(5):98-105.
[20]沈苑,胡梦圆等.可信赖人工智能教育应用的建设路径与现实启示——以英国典型举措为例[J].现代远程教育研究,2023,35(4):65-74.
[21]王佑镁,王旦等.算法公平:教育人工智能算法偏见的逻辑与治理[J].开放教育研究,2023,29(5):37-46.
[22]Zhang C,Xie Y,et al.A Survey on Federated Learning[J].Knowledge-Based Systems,2021,216:106775.
[23]Dang M,Wang H X,et al.Explainable Artificial Intelligence:A Comprehensive Review[J].Artificial Intelligence Review,2022,55(5):3503-3568.
[24]纪守领,李进锋等.机器学习模型可解释性方法、应用与安全研究综述[J].计算机研究与发展,2019,56(10):2071-2096.
[25]Sawyer,R.Keith,ed.The Cambridge Handbook of the Learning Sciences[M].Cambridge:Cambridge University Press,2005.
[26]卢宇,骈扬等.新型智能导学系统构建及其关键技术[J].中国远程教育,2023,43(7):30-38+46.
[27]戴岭,祝智庭.教育人工智能伦理与道德风险治理:问题廓清与精准施策[J].中国教育学刊,2024,(12):31-37.
[28]Li L,Fan Y,et al.A Review of Applications in Federated Learning[J].Computers&Industrial Engineering,2020,149:106854.
[29]马安瑶.基于联邦学习的可信知识追踪模型构建研究[D].北京:北京师范大学,2025.
[30]Falmagne J C, Koppen M, et al.Introduction to Knowledge Spaces:How to Build,Test,and Search Them[J].Psychological Review,1990,97(2):201-224.
[31]Murre J M J,Dros J.Replication and analysis of Ebbinghaus’forgetting curve[J].PLo S one,2015,10(7):e0120644.
[32]Lu Y,Wang D,et al.Design and Evaluation of Trustworthy Knowledge Tracing Model for Intelligent Tutoring System[J].IEEE Transactions on Learning Technologies(TLT),2024,17:1661-1676.
基本信息:
中图分类号:G434
引用信息:
[1]卢宇,雷奕欣,陈鹏鹤.可信教育人工智能模型构建与应用研究[J].中国电化教育,2026,No.468(01):38-45.
基金信息:
国家自然科学基金面上项目“可信人工智能驱动的知识追踪模型构建研究”(项目编号:62477003)研究成果
2026-01-10
2026-01-10