面向学习云空间的认知投入量化研究

Quantifying cognitive engagement in cloud-based learning space

在全面推进"网络学习空间人人通"的建设背景下,学习云空间的应用不断深入。但正如其他在线学习平台一样,学习云空间也存在辍学率高、学习投入不足等问题。本文聚焦学习云空间中学习主体认知投入相关数据的获取,提出了基于机器学习的认知投入量化方法。首先,分析了认知投入的构成要素,并基于班杜拉的社会认知理论,建立了面向学习云空间的认知投入模型。然后构建了围绕"数据采集-数据处理-量化实现-量化应用"的认知投入量化框架,并设计了一个基于支持向量机的认知投入量化算法。最后以世界大学城系统平台数据为支撑,通过与其他机器学习算法的对比得到研究结果,基于SVM算法具有较高的量化精确率,本研究希望能为教育领域深层认知投入量化提供一种可行的参考方案。

Despite the tendency that cloud-based learning platforms are in increasing use against the backdrop of ensuring online learning space for all,high dropout rates and insufficient learning engagement are not uncommon,as is the case with other online learning platforms.This study sets out to measure cognitive engagement in cloud-based learning space.It first analyzes the components of cognitive engagement and draws upon Bandura’s social cognitive theory to build a model for cognitive engagement in cloud-based learning space.A quantification framework of data collection,data processing,quantification and application is then created,supported by a Support Vector Machine algorithm to measure cognitive engagement.This SVM algorithm is found to be more accurate than other algorithms when applied to data from WorldUC platform.