推进多模态大模型的教育深度应用:价值赋能、场景构建与实施建议  

Advancing the Deep Application of Multimodal Large Models in Education:Value Empowerment,Scenarios Construction,and Implementation Strategies

作　　者：杨现民[1] 卜浩德 李新[1] YANG Xianmin;BU Haode;LI Xin

机构地区：[1]江苏师范大学江苏省教育信息化工程技术研究中心,江苏徐州221116

出　　处：《中国教育学刊》2025年第4期9-14,共6页Journal of The Chinese Society of Education

基　　金：江苏省高等教育教改研究重中之重课题“基于‘5G+VR’的师范生远程沉浸式课堂教学实训体系构建与创新实践”(项目编号:2023JSJG018)阶段性研究成果。

摘　　要：以多模态大模型为核心的人工智能技术正在重塑社会组织与运行模式,成为驱动社会结构变革和文化演变的重要力量。多模态大模型具备多模态内容理解、多模态资源生成、多模态人机交互等技术特征,其教育应用价值体现在助力多模态学习分析、优化数字资源供给方式以及赋能智适应学习三个方面,典型应用场景涵盖教学辅助、学习辅导、教育管理、教育评价以及教育研究。为进一步推进多模态大模型在教育领域的深度应用,研究提出五条实施建议:明确应用需求,拓展多模态大模型教育应用场景;注重素养提升,强化教育利益共同体多模态大模型应用能力;坚持三项原则,提升多模态大模型与教学深度融合的效度;选好技术产品,建立多模态大模型教育服务支持体系;关注应用绩效,持续开展多模态大模型教育影响研究。The artificial intelligence technology with multimodal large model as its core is reshaping the society's organizing and operating modes,becoming a key force driving the change of social structure and cultural evolution.Multimodal large models are characterized by technologies such as multimodal content understanding,multimodal resource generation,and multimodal human-computer interaction.Their educational application value is reflected in three aspects:supporting multimodal learning analytics,optimizing digital resource supply methods,and empowering intelligent adaptive learning.Typical application scenarios include teaching assistance,learning tutoring,educational management,educational evaluation,and educational research.To further promote the deep application of multimodal large models in the field of education,this study proposes five implementation suggestions:clarify the application needs and expand the application scenarios of multimodal large models in education;focus on enhancing literacy and strengthening the multimodal large model application capabilities of the educational community;adhere to three principles to improve the effectiveness of the deep integration of multimodal large models into teaching;carefully select technical products and establish a multimodal large model education service support system;and focus on application performance and continuously carry out research on the educational impact of multimodal large models.

关 键 词：多模态大模型 教育应用 价值赋能 场景构建 实施建议 

分 类 号：G434[文化科学—教育学]