基于教育大数据的学习干预模型构建  

The Construction of a Learning Intervention Model Based on Big Data Analytics

作　　者：李彤彤[1] 黄洛颖 邹蕊[3] 武法提[3] 

机构地区：[1]天津师范大学教育科学学院,天津300387 [2]北京师范大学信息网络中心,北京100875 [3]北京师范大学教育学部教育技术学院,北京100875

出　　处：《中国电化教育》2016年第6期16-20,共5页China Educational Technology

基　　金：2014年全国教育科学"十二五"规划教育部重点课题"基于教育大数据的学习分析工具设计与应用研究"(课题编号:DCA140230)研究成果

摘　　要：学习分析是利用预测模型及学习过程中产生的数据,分析、预测学习者未来表现并发现潜在问题以实施干预的技术。在学习分析过程中,学习干预作为与教学过程直接相联的部分,是改善、提升学习成效的关键。然而,当前国内外已有的学习干预方面的相关研究大多从微观层面上列举了部分干预措施,缺乏整体性和系统性的考虑。该研究构建了基于教育大数据和学习分析的、以干预引擎为核心的"状态识别—策略匹配—干预实施—成效分析"四环节循环结构干预模型,该模型主要关注学习者的学习风格类型、学习进度水平、学习互动水平、学业成就水平四方面的状态特征,并针对这四方面的状态水平设计了具体的干预策略、干预时机以及干预方式,以期为在线学习环境中的学习干预理论研究与实践应用提供参考。Learning analytics aims to analyze and predict learners＇ future performance, find potential problems and implement intervention according to the prediction model and big data generated in the learning process. As it is attached directly to teaching process, learning intervention is the key to improve learners＇ learning performance. However, most of existing researches on learning intervention propose intervention measures from micro-level, lack of systemic and integral consideration. This paper constructs a learning intervention model based on learners＇ big data analytics, which is a loop structure with intervention engine as its core. The intervention model consists of four parts： learners＇ status identification, intervention strategy matching, intervention implementation, and intervention effectiveness analysis. As to learners＇ status, the model mainly focus on learning style, learning progress, interaction level, and learning performance. We also propose the typical intervention strategies, intervention occasion, and specific intervention means according to learners＇ status level in this model. The significance of this model is that it can provide reference for academic research and practical application of learning intervention in online learning environments.

关 键 词：干预模型 干预策略 学习分析 教育大数据 

分 类 号：G434[文化科学—教育学]