【作 者】：
王洪江， 张一夫， 伦 昊， 陈沛瑜， 张少英
【关 键 词 】：
在线学习元认知能力； 数字化模型； 多模态评估； 常态化教学； 深度学习
【栏 目】：
网络教育
【中文摘要】：
在线学习元认知能力对在线学习成效具有重要影响，对其进行模型构建有助于学习者调整学习策略和过程，但当前建模方式存在理论指导不统一、偏向于单模态指标等困境。为此，文章梳理在线学习元认知理论，构建基于多模态数据的在线学习元认知能力数字化模型，并将其应用于常态化在线课程中，以验证基于模型的在线学习元认知能力评估效果。结果表明：（1）基于多模态数据的数字化模型为准确评估在线学习元认知能力提供了基础；（2）各模态数据采用深度学习模型进行分析，结合多模态决策级融合，使评估结果具有全面性与可解释性。
【英文摘要】：
Online learning metacognitive ability has an important impact on the effectiveness of online learning, and modeling it helps learners to adjust their learning strategies and processes. However, current modeling approaches suffer from inconsistent theoretical guidance and an over-reliance on unimodal indicators. For this reason, the study combs through online learning metacognition theories, constructs a digital model of online learning metacognitive ability based on multimodal data, and applies it to regular online courses to verify the effectiveness of model-based assessment of online learning metacognitive ability. The results demonstrate that: (1)the multimodal data-based digital model provides a solid foundation for accurately assessing online learning metacognitive ability; (2)each modal data is analyzed using a deep learning model combined with multimodal decision-level fusion to make the assessment results comprehensive and interpretable.