机器学习领域的研究热点及主题演化分析——以科学计量为基础

Research Hot Spots and Topic Evolution Analysis in the Field of Machine Learning——Based on the Scientific Measurement

以WOS核心数据集为平台,以机器学习为主题,采集2008年1月—2019年7月发表的论文,经过预处理后共获得7279条文献记录。借助信息计量工具、文本挖掘技术和知识图谱可视化工具,对机器学习领域的研究成果进行年代分布、发文地区及机构、关键词共现、研究前沿与热点等情况进行综合分析,并以图表方式对机器学习研究的重点问题、领域、趋势进行总结。研究发现:机器学习的研究大多集中于支持向量机、分类和预测、神经网络、人工智能、深度学习等领域,下一阶段将会在量子计算、算法改进、相变等领域展开新一轮研究热潮。

Using WOS core data set as the platform and machine learning as the theme,this article collects the papers published from January 2008 to July 2019 and obtains a total of 7279 literature records after preprocessing.With the help of information measurement tools,text mining technology and knowledge map visualization tools,it comprehensively analyzes the research results in the field of machine learning according to chronology distribution,publishing areas and institutions,keyword co-occurrence,research frontiers and hot spots,etc.It summarizes the key issues,fields and trends of machine learning research in the form of diagrams.It finds that most of the research on machine learning focuses on support vector machine,classification and prediction,neural networks,artificial intelligence,deep learning and other fields.In the next stage,a new wave of research will be carried out in quantum computing,algorithm improvement,phase transition and other fields.