基于LDA与DTM模型的粤港澳大湾区文献主题演化研究

Research on the Evolution of Literature Topics in the Guangdong-Hong Kong-Macao Greater Bay Area Based on LDA and DTM Models

利用中国知网(CNKI)数据,基于CiteSpace可视化文献分布情况,采用隐含狄利克雷分布(LDA)主题模型提取关键词,挖掘出文本隐含的研究主题及重要性程度;结合动态主题模型(DTM),分析不同主题内容的演化过程。结果表明,粤港澳相关研究主题主要集中于经济合作、协同治理、人才培养、环境保护、产业升级、旅游与文化六大方面。其中,经济合作和协同治理是目前研究关注的重点,人才培养和环境保护是今后主要研究趋势,且相关内容呈现不断细化特征。

This paper uses CNKI data,based on Citespace to visualize the distribution of documents,and uses the LDA topic model to extract keywords to find out the hidden research topics and importance of the text;combined with the DTM topic model,the evolution process of different topics is analyzed.The results show that the research topics related to Guangdong,Hong Kong and Macao mainly focus on six aspects:economic cooperation,collaborative governance,personnel training,environmental protection,industrial upgrading,tourism and culture.Among them,economic cooperation and collaborative governance are the focus of current research,while personnel training and environmental protection will be the main research trends in the future,and the relevant content will continue to be refined.