远程教育杂志2026，Vol.44Issue(3)：72-82,11.DOI:10.15881/j.cnki.cn33-1304/g4.2026.03.008
生成式教育评价:混合增强智能驱动的人机协同评价范式创新与实践
Generative Educational Evaluation:Paradigm Innovation and Practice of Human-Machine Collaborative Evaluation Driven by Hybrid-augmented Intelligence
摘要
Abstract
The paradigm of educational evaluation has continued to evolve alongside technological advancement.The rise of generative artificial intelligence is driving a shift from perceptual intelligence to cognitive intelligence,thereby creating new techno-logical opportunities for upgrading intelligent educational evaluation.However,existing human-machine collaborative evaluation ap-proaches still lack an end-to-end intelligent implementation mechanism.Hybrid-augmented intelligence based on large language models has key capabilities such as multi-source knowledge integration,complex activity understanding,and human-machine collab-orative verification.These capabilities provide technological support for theoretical innovation in educational evaluation and offer practical pathways for evaluation reform.Guided by the theory of hybrid-augmented intelligence and grounded in the value orientation of talent cultivation,this study clarifies the logical framework of generative educational evaluation and develops a generative educa-tional evaluation model supported by human-machine collaborative strategies.It further explains the technical mechanism through which generative educational evaluation becomes both"computable"and"decision-supportive,"including six structural modules:holographic data collection,multidimensional data analysis,automatic indicator discovery,automatic feedback generation,educational evaluation optimization,and trustworthy recommendation for decision-making.Based on this framework,the study explores the appli-cation of generative educational evaluation in multiple scenarios,including student academic assessment,teachers' instructional en-gagement evaluation,and university program monitoring.The findings indicate that a full-process,multi-granularity educational sys-tem of data collection and governance,prompting strategies that integrate value orientation with professional knowledge,and a multi-level validity-based collaborative verification mechanism provide critical support for translating generative educational evaluation from theory into practice.Overall,generative educational evaluation integrates the complementary strengths of human cognitive wisdom and machine-based data perception.Through holographic and multidimensional measurement of educational processes and outcomes,it can iteratively generate personalized value judgments and provide a new paradigm for educational evaluation and decision-making.关键词
生成式教育评价/教育大模型/混合增强智能/生成式人工智能/智能评价/人机协同Key words
Generative educational evaluation/Educational large language models/Hybrid-augmented intelligence/Generative artificial intelligence/Intelligent evaluation/Human-machine collaboration分类
社会科学引用本文复制引用
熊余,蔡婷,肖春玲,袁春艳..生成式教育评价:混合增强智能驱动的人机协同评价范式创新与实践[J].远程教育杂志,2026,44(3):72-82,11.基金项目
国家自然科学基金面上项目"教师课堂教学投入的智能识别与可解释评价研究"(项目编号:62377007)、重庆市教育科学规划重点课题"大模型赋能重庆高等教育创新发展的对策研究"(项目编号:K25YD2060056)、重庆市教委科学技术研究重大项目"人机共生学习环境下可解释学习推荐技术研究"(项目编号:KJZD-M202400606)、重庆市教委科学技术研究青年项目"基于情境感知的视频表示学习及其在教师授课行为理解中的应用研究"(项目编号:KJQN202400634)、成都市区域科技创新合作项目"多模态大模型驱动的智慧课堂教学行为分析与生成式评价"(项目编号:2026-YF11-00042-HZ). （项目编号:62377007）
