【作 者】：
徐振国， 王佳宁， 王 悦， 郭顺利， 谢万里
【关 键 词 】：
在线课程； 自媒体平台； 课程质量； 深度学习； 课程评价
【栏 目】：
网络教育
【中文摘要】：
自媒体平台存在海量在线课程，但质量良莠不齐，严重影响学习者的学习体验和学习效率，因此，探索客观、准确、高效的自媒体平台在线课程质量评价方法成为亟待解决的现实问题。针对该问题，研究首先构建在线课程质量评价体系，并利用层次分析法确定各指标权重。其次，利用深度学习和自然语言处理等技术对在线课程评论数据进行粗粒度情感分析和细粒度情感分析，以实现自媒体平台在线课程质量评价。最后，采集Bilibili平台在线课程评论数据进行应用研究，并通过实验证实该方法具有较高的可行性、可信度和准确率，可实现对自媒体平台在线课程质量的有效评价，以促进在线课程良性发展，优化学习者的学习体验。
【英文摘要】：
There are a large number of online courses on we media platforms, but the quality is uneven, which seriously affects the learning experience and learning efficiency of learners. Therefore, exploring objective, accurate and efficient quality evaluation methods of online courses for we media platform has become a practical problem to be solved. To address this problem, this study firstly constructs an online course quality evaluation system, and uses the analytic hierarchy process to determine the weight of each index. Secondly, deep learning and natural language processing technologies are used to conduct coarse-grained sentiment analysis and fine-grained sentiment analysis of online course review data in order to realize the quality evaluation of online courses on we media platforms. Finally, the review data of online courses on Bilibili platform is collected for application research, and the experiment has proved that the method has high feasibility, credibility and accuracy, and can realize the effective quality evaluation of online courses on we media platforms, so as to promote the healthy development of online courses and improve the learning experience of learners.