面向MOOC课程评论的学习者话题挖掘研究

发布时间：2018-04-11 12:47

本文选题：MOOC + 文本评论　；参考：《电化教育研究》2017年10期

【摘要】：研究以果壳网MOOC学院的"财务分析与决策"课程为实验对象,通过分析课程评论帖进行学习者话题的挖掘。文章不仅采用了高频词汇分析的定量方法,实现对学习者课程评论内容的整体认识,并且,根据参与评论学习者的课程完成情况,分别对已完成和未完成两种类型的学习者展开定性的学习分析研究,应用非监督学习方法 LDA模型自动挖掘和解析文本评论信息的特征结构和语义内容,并探究和追踪学习者关注的热点话题演化趋势。实验结果表明,学习者认可和赞赏了该门课程,并且尤为关注课程内容以及教师授课形式话题;相比课程完成者,未完成者更倾向于解释其未完成课程的主要原因,表达出更为消极的话题内容,并较少涉及课程本身相关的专业理论知识。
[Abstract]:Taking the course of Financial Analysis and decision making in MOOC College of Fruit Shell as the experimental object, the topic mining of learners was carried out by analyzing the comments of the course.This paper not only uses the quantitative method of high-frequency lexical analysis to realize the overall understanding of the content of the learner's curriculum review, but also, according to the completion of the course,Qualitative learning analysis is carried out on two types of learners: completed and uncompleted, and LDA model is used to automatically mine and analyze the feature structure and semantic content of text comment information.And to explore and track the evolution trend of hot topics concerned by learners.The experimental results show that the learners recognize and appreciate the course, and pay particular attention to the topic of the content of the course and the teaching form of the teacher, and that the incomplete students are more likely to explain the main reasons for the unfinished course than the ones who have completed the course.To express more negative topics, and less related to the curriculum itself professional theoretical knowledge.
【作者单位】：华中师范大学国家数字化学习工程技术研究中心;
【基金】：2016年教育部—中国移动科研基金项目“国家教育大数据相关问题研究”(项目编号:MCM20160401);2016年度教育部人文社会科学研究青年基金项目“高校慕课环境下的互动话语行为及其对学习效果的影响机理研究”(项目编号:16YJC880052)
【分类号】：G434

【相似文献】