数据驱动的在线学习表现模型构建研究

Research on Construction of Data Driven Online Learning Performance Model

随着MOOC的兴起,在线学习已经普及并得到认可。如何快速而准确地评价学习者在线学习表现,成为大家关注的问题。大部分或绝大部分的课程的学习行为都发生在学习平台上,这使得基于MOOC平台的在线学习行为日志数据来描述学习者在线学习表现是有效的。那么,从哪些方面来描述在线学习表现,成为该文所关注的问题。该文选择Moodle平台为例,以一门在线课程为样本,基于Moodle在线学习行为日志数据构建在线学习绩效模型。该模型基于在线学习行为日志数据动态生成,不同课程会有所不同。实践应用表明,该模型可以较好地描绘学习者的在线学习表现,成为快速而准确地自动评价学习者的基础。

With the rise of MOOC, online learning has been widely recognized. how to make a quick and accurate assessment of learners＆#39; online learning performance becomes the focus of attention. With the support of MOOC, most of the learning behaviors of the course happen on the learning platform, which makes describing online learning behavior based log data of MOOC platform is effective. Therefore, which aspects should be used to describe online learning performance, becomes this problem that the paper is concerned. This paper chooses Moodle platform as an example, takes a pure online course as the sample, and build an online learning performance model based on Moodle log data. The model is dynamically generated based on the online learning behavior log data of which different courses could be different. The practical application shows that the model can well describe the online learning performance of learners, and it is the basis of a fast and accurate automatic evaluation of learners.