摘 要

Abstract 2

1 绪论 3

1.1 研究背景 3

1.2 研究意义 4

1.2.1 理论意义 4

1.2.2 实际意义 4

1.3 国内外发展及研究现状 4

1.3.1 Mooc平台的发展现状 5

1.3.2 国内外研究现状 5

1.4 研究思路及方法 6

2 相关理论概述 9

2.1 学习者满意度 9

2.2 文本分析 10

2.3 LDA话题模型 10

3 实验过程及总结 12

3.1数据的获取 12

3.2 数据的预处理 12

3.2.1 空值和异常值的处理 12

3.2.1 补充用户自定义分词字典 12

3.3 实验分析 12

3.3.1 高频词分析 13

3.3.2 话题挖掘 16

3.4 实验结果 24

4 总结和展望 25

4.1 研究结论 25

4.2 建议 25

4.2 研究不足和展望 26

4.2.1 研究不足 26

4.2.2 研究展望 27

参考文献 28

致 谢 29

摘 要

近年来，由于远程教育的不断发展，新媒体技术的不断更新，网络信息技术的不断进步，大规模开放网络在线课程（Mooc）风靡全世界。由于其优质、开放、免费等特征，吸引了许多学习者来使用，对在线教育领域产生了深刻的影响。许多高校争相建立Mooc平台并与其他高校合作，拥有许多在线学习者。但是，根据网上调查显示，Mooc的学习效果并不是十分理想，甚至有些课程的完成率不足5%。对于Mooc平台来说，学习者是Mooc的服务对象，也是Mooc的直接使用者。所以，围绕学习者开展相关研究以提高学习者对Mooc平台的满意度就是当前完善Mooc的重要课题。课程评价作为学习者对课程产生的直接文本，其中包含着学习者的学习过程和学习体验等数据，同时也可以通过评分体现出学习者对课程的满意度。

本研究通过java爬虫爬取中国Mooc平台课程的评论，累积评论74873条,对评论文本进行预处理后有效数据74256条（筛除省略号、emoji表情、符号乱码等无用评论）。然后统计评论文本中的高频词生成词云图，对评论文本有整体的概括和把握，再对评论文本建立话题模型，并基于不同课程类型、最热课程、最高评价课程和最低评价课程对比等多个维度来分析话题模型，最终分析出影响Mooc学习者满意度的主要因素为教学效果、教学方法、教学内容、课程资源，对国内MOOC平台的改善、网络课程的优化提供有力的理论支持，为学习者真正的享受学习、体现 MOOC 的价值以及实现教育与技术的深度融合做出贡献。

关键词：中国大学Mooc平台；学习者满意度；影响因素；文本分析

Abstract

In recent years, due to the continuous development of distance education, the continuous updating of new media technologies, and the continuous advancement of network information technology, large-scale open online courses (Mooc) have swept the world. Because of its high quality, openness, and free features, it has attracted many learners to use and has had a profound impact on the field of online education. Many colleges and universities compete to establish the Mooc platform and cooperate with other universities and have many online learners. However, according to online surveys, Mooc's learning results are not very satisfactory, and even some courses have a completion rate of less than 5%. For the Mooc platform, learners are Mooc's service objects and Mooc's direct users. Therefore, it is an important topic for improving Mooc to conduct relevant research around learners to improve learners' satisfaction with the Mooc platform. As a direct text produced by the learner, the course evaluation contains the learner's learning process and the learning experience. At the same time, the course evaluation can reflect the learner's satisfaction with the course.

In this study, the java reptile climbed through the reviews of the Chinese Mooc platform curriculum, accumulating 74,873 comments, and pre-processing the review text with 74,256 valid data (elimination of useless comments such as ellipses, emoji expressions, and symbol garbled). Then, the high-frequency words in the commentary texts are generated by the word cloud map, and the overall summary and grasp of the commentary texts. Then the topic model is established for the commentary texts, and based on different course types, the hottest courses, the highest evaluation courses and the comparison of the lowest evaluation courses, etc. Multiple dimensions are used to analyze the topic model. Finally, the main factors influencing Mooc learner satisfaction are the teaching effects, teaching methods, teaching contents, and curriculum resources. They provide strong theoretical support for the improvement of the domestic MOOC platform and optimization of online courses. It contributes to the true enjoyment of learning, the value of MOOCs, and the deep integration of education and technology.

Key Words：Chinese University Mooc Platform; Learner Satisfaction; Influencing Factors; Text Analysis

1 绪论

1.1 研究背景

近年由于远程教育、新媒体技术、网络信息技术的不断发展，大规模网络课程作为一种新型的在线教育模式诞生了，其开放和免费等特性为其吸引了大量的学习者。而且MOOC对学习者的碎片化学习提供了强有力的支持，其打破了时、空间的特性使学习者可以在任何地点、任何时间进行学习。2012年美国，MOOC开始大规模发展，各大名校均推出相应的免费课程，Coursera、Udacity、edX三大课程提供商的兴起，给更多学生提供了系统学习的可能。后来掀起了MOOC发展的潮流，全世界其他顶尖大学也开始提供大型开放式网络课程，将网络的便捷和教育相结合，通过网络传播学习信息，使每个人都可以享受顶尖大学的教育资源，并且可以随时随地进行学习，这就是MOOC带来的重大影响。