基于教育大数据的大规模私有在线课程中交互活动模式的研究

A Study of Interaction Activity Patterns in Massive Private Online Courses Based on Educational Big Data

作为我国开放高等教育的主要形式,大规模私有在线课程(MPOCs)的学习管理系统积累了大量学习者行为数据。对于这些数据的分析,能够有效地探究教学与学习的规律,提高开放高等教育的实践和评估。文章运用隐马尔科夫模型识别了MPOCs中不同课程阶段的学习者交互活动的序列模式。研究结果表明,MPOCs中的学习者在学习周和考试周的行为模式存在明显差异,他们会根据不同的课程阶段调整学习策略和侧重点,但小组学习和讨论对于学习成效的影响不大,这一研究结果与传统主流研究中关于协作学习及讨论的正向研究结果有比较明显的差异。这些研究结果有助于学习者和教师动态地梳理教育教学过程,促使他们采用不同学习策略来促进教学与学习,以及从更加细致的角度来评估开放高等教育的教学效果。

As the main form of open higher education in China,the learning management system of massive private online courses(MPOCs)has accumulated a large amount of learner behavior data.The analysis of these data can effectively explore the patterns of teaching and learning and improve the practice and evaluation of open higher education.This paper uses the Hidden Markov model to identify the sequential patterns of learner interaction activities at different course stages in MPOCs.The results indicate that the learners in MPOCs differ significantly in their behavioral patterns during study weeks and examination weeks.They can adjust their learning strategies and focus according to different course phases,but group learning and discussion have little effect on learning effectiveness.This result is totally different from the positive results of traditional mainstream research on cooperative learning and discussion.These findings help learners and teachers to sort out the educational and teaching process dynamically in the teaching and learning process,promote them to adopt different learning strategies to facilitate teaching and learning,as well as to evaluate the teaching and learning of open higher education from a more detailed perspective.