【作 者】：
田 浩， 武法提
【关 键 词 】：
复杂性科学； 学习干预； 智能教育； 模型构建
【栏 目】：
理论探讨
【中文摘要】：
迅猛发展的智能技术在变革教育样态的同时，也使得教育的复杂性本质日益凸显。立足复杂性科学视角开展精准化、个性化的学习干预有利于满足学习者学习实践中的现实需求。文章在阐述了复杂性科学与学习干预的基本概念之后，对复杂性科学指导学习干预实施的适切性进行了分析。进而从学习问题诊断、干预策略匹配、干预策略实施、干预结果分析四个核心要素出发，分别解析了学习干预的复杂性特征，并由此构建了复杂性科学视域下的学习干预模型，包含问题诊断层、动力引擎层、推理匹配层和进化适应层四个逻辑层次，依次描绘了基于学习分析技术的学习问题发现、基于元素自组织的学习问题归因、基于因果链推理的干预策略匹配以及基于多主体进化的干预策略实施。该模型形成了完整的智能学习服务路径，为智能时代开展人机协同的精准学习干预提供了理论与实践依据。
【英文摘要】：
The rapid development of intelligent technology has not only transformed the style of education, but also brought the complexity of education to the fore. Precise and personalized learning interventions based on the perspective of complexity science are conducive to meeting the real needs of learners in learning practice. After explaining the basic concepts of complexity science and learning interventions, this paper analyzes the applicability of complexity science to guide the implementation of learning interventions. Then, this paper analyzes the complexity characteristics of learning interventions in terms of four core elements of learning problem diagnosis, intervention strategy matching, intervention strategy implementation, and intervention effect analysis. A learning intervention model is constructed from the perspective of complexity science, which consists of four logical levels: problem diagnosis, motivation engine, reasoning matching, and evolutionary adaptation. It successively depicts the discovery of learning problems based on learning analytics technology, the attribution of learning problems based on elemental self-organization, the matching of intervention strategy based on causal chain reasoning, and the implementation of intervention strategy based on multi-agent evolution. This model forms a complete intelligent learning service path, which provides a theoretical and practical basis for carrying out precise learning interventions of human-computer collaboration in the intelligent era.