基于TF-IDF算法和DTM模型的网络学习社区主题分析  

Topic Evolution Characteristics in Domestic E-learning Community Based on TF-IDF Algorithm and DTM

作　　者：董伟 董思遥 王聪 陶金虎 DONG Wei;DONG Si-yao;WANG Cong;TAO Jin-hu(School of Education,Tianjin University,Tianjin,China 300350)

机构地区：[1]天津大学教育学院,天津300350

出　　处：《现代教育技术》2022年第2期90-98,共9页Modern Educational Technology

基　　金：2020年天津市教委社科重大课题“混合教学环境下课程资源建设与教学模式优化对策研究”(项目编号:2020JWZD03)的阶段性研究成果。

摘　　要：随着信息技术的快速发展,网络学习社区已成为重要的学习平台之一。在此背景下,文章采用词频—逆文本频率(Term Frequency-Inverse Document Frequency,TF-IDF)算法和动态主题模型(Dynamic Topic Models,DTM),以951条相关文献作为数据来源,首先对网络学习社区的研究热点和研究主题进行了分析,发现网络学习社区具有明确的教育属性,并将网络学习社区领域的相关研究划分为九类主题。随后,文章基于时序绘制了热度高低均值图和热度曲线图,对网络学习社区主题热度与演化情况进行了分析,发现混合教学和学习服务等主题具有核心热点特征,平台建设、学习者交互、社区生态等主题发展较为稳定,学习环境主题有一定的发展潜力,而学习模式、知识共享、学习动机等主题热度逐渐弱化。最后,文章梳理了研究结论,并针对网络学习社区的发展提出了建议。文章的研究为网络学习社区领域中理论和实践研究提供了一定的数据和理论支持,明确了该领域的研究主题和结构,为探讨该领域的未来研究方向提供了新的思路。With the rapid development of information technology,the E-learning community has become one of the important learning platforms.Under this context,this paper uses the word frequency-inverse text frequency(TF-IDF)algorithm and dynamic topic model(DTM) and takes 951 relevant documents as data sources.First,this paper analyzes the research hotspots and topics of the E-learning community,finds out that the E-learning community has the clear attribute of education,and divides the relevant research in this field into 9 topics.Then,this paper draws the heat high-low mean map and heat curve to analyzes the theme heat and evolution map of e-learning community based on time series.It finds that the mean heat of mixed teaching and learning service is relatively high,which is a theme with core hot features.The topics such as platform construction,learner interaction,and community ecology are relatively stable in their development.The learning environment has great development potential.The heat of topics such as learning mode,knowledge sharing,and learning motivation is gradually weakening.Finally,this paper comes to a conclusion and gives some suggestions for the development of E-learning community.The research results provide some data and theoretical support for the theoretical and practical research in the field of the e-learning community,define the research theme and structure of this field,and provide new ideas for exploring the future research direction of this field.

关 键 词：网络学习社区 TF-IDF算法 DTM模型 主题热度 演化情况 

分 类 号：G40-057[文化科学—教育学原理]