【作 者】：
牟智佳，岳 婷，朱 陶
【关 键 词 】：
人机协同； 认知智能大模型； 个性化学习； 学习设计； 实践模式
【栏 目】：
课程与教学
【中文摘要】：
探索规模化教育与个性化培养有机结合是新时代教育改革发展的一个重要学术命题。以大语言模型为代表的生成式人工智能改变了知识获取和人机交互方式，为提供定制化学习体验和创新个性化学习设计带来了新的驱动力。人机协同视域下的个性化学习设计在转变理路上呈现出由片段式向全程化、由推送式向定制化、由单模态向多模态的本体流变，在发展理路上表现为人机共教、人机共学、人机共评、人机共管的多重向度，在逻辑理路上体现为导引目标、激活进程、定制方法、创新范式的适配赋能。个性化学习设计的结构体系是以学习者差异认知基础为逻辑起点，以支持学习者个性发展为目标导向，以人机协同对话提示语模式为基础，通过学习模块设计与协同路径规划构建个性化学习流程框架，形成三元主体协同、适性服务闭环的个性化学习设计范型。人机协同对话提示语模式依托内容生成机制与提示工程技术，通过需求描述、角色设定、示例指定引导对话轮转以促进个体思维链形成，为个性化学习过程提供协同支持。研究探讨以类人化智慧赋能个性化学习交互场景形成新的人机协同学习样态，为垂类模型的教育应用提供有益参考。
【英文摘要】：
Exploring the organic combination of large-scale education and personalized cultivation is an important academic proposition for educational reform and development in the new era. The generative artificial intelligence represented by large language models has changed the way of knowledge acquisition and human-computer interaction, and brought a new drive to provide customized learning experience and innovative personalized learning design. The personalized learning design in the context of human-computer collaboration presents an ontological change from the fragmented to the holistic, from the push-based to the customized, and from the unimodal to the multimodal in the process of transformation. In the process of development, it is embodied in the multiple dimensions of human-computer co-teaching, human-computer co-learning, human-computer co-evaluation, and human- computer co-management. In the process of logic, it is manifested in the adaptive empowerment of guiding goals, activating processes, customizing methods, and innovating paradigms. The structural system of personalized learning design takes the different cognitive foundation of learners as the logical starting point, and aims to support the development of learners' personality. Based on the human-computer collaborative dialogue prompt pattern, a framework of personalized learning process is constructed through the design of learning modules and the planning of collaborative paths, forming a personalized learning design paradigm with three-dimensional collaboration of the main body and a closed-loop of adaptive services. The human-computer collaborative dialogue prompt pattern relies on the content generation mechanism and prompt engineering technology, and guides dialogue turns through demand description, role setting, and example specification to promote the formation of individual thought chains, providing collaborative support for personalized learning processes. The study explores the formation of a new human-machine collaborative learning pattern empowered by humanoid intelligence for personalized learning interaction scenarios, and provides a useful reference for the educational application of the vertical model.