我国深度学习研究的演进阶段及热点趋势分析——基于教育技术八种核心期刊论文的可视化分析

The Analysis of the Evolution Phase and Hotspot Trend of Deep Learning Research in China—Based on the Visualized Analysis of Eight Core Journal Papers of Education and Technology

本文以教育技术八种核心期刊中与深度学习主题相关的214篇论文为研究对象,在利用COOC、SPSS等软件对文献数据进行高频关键词词频分析、时区分布分析、聚类分析及多维尺度分析的基础上绘制知识图谱。根据高频关键词时区分布及文献梳理情况,将国内深度学习研究划分为依托学习科学基本研究的萌芽阶段、融入多种学习方式及网络技术的发展阶段、扎根智慧教育的成熟阶段三个阶段。关键词聚类树状图及多维尺度图谱结果表明,当前我国深度学习研究主要集中在技术支持、教学模式、学习活动以及应用领域四个方面。展望未来,技术支持的深度学习研究常态化、不同教学模式深度学习影响因素分析、深度学习研究在学习科学分支中进一步细化以及拓宽深度学习研究的学科融合度将是我国深度学习研究的重点。

This paper regards 214 papers related to the theme of deep learning in the eight core journal papers of education and technology as research object,and utilizes software like COOC,SPSS,etc.to make analysis of frequency of key words,time zone distribution analysis,cluster analysis,and multidimensional scale analysis,based on which knowledge map is drawn.According to the condition of time zone distribution of high-frequent key words and literature review,this research has divided the research of deep learning into three phases,namely,the budding phase relying on the basic research of learning science,the developing phase of integrating many kinds of learning styles and network technology,and mature phase being rooted in smart education.The results of key words cluster tree map and multidimensional scale maps show that the present national research of deep learning mainly focuses on four aspects,namely,the technical support,teaching mode,learning activity,and application field of deep learning.Looking forward to the future,the key points of the research on deep learning in China will be the normalization of deep learning research of technical support,the analysis of the influential factors of deep learning of different teaching modes,the further refinement of deep learning research in the branch of learning science,and the integration of disciplines of broadening deep learning research.