科技广场Issue(5)：16-29,14.
基于LDA主题模型的我国氢能技术融合发展趋势分析
Analysis of Technology Convergence Trends in China's Hydrogen Energy Sector Based on LDA Topic Model
摘要
Abstract
Based on 283 national and local hydrogen industry policy documents issued from 2022 to 2024,this study employs the Latent Dirichlet Allocation(LDA)Topic Mod-el to identify six core themes:hydrogen production technology,storage and transportation technology,fuel cells,application scenarios,infrastructure,and industrial policies.Then it makes a statistical analysis of such indicators as characteristic term frequency,theme attention,technology convergence intensity,and Pearson correlation coefficient to examine policy evolution,theme popularity,and technology convergence characteristics.The findings are as follows:Policy theme attention demonstrates phased characteristics.Industrial policy,despite a year-on-year decline,remains the most prominent.The popularity of hydrogen production technology and infrastructure witnessed a notable drop in 2024,while storage and transportation technology remained the least concerned.There are significant differences in the convergence intensity between themes.Hydrogen produc-tion technology and storage/transportation technology exhibit the strongest synergy,whereas hydrogen production technology and application scenarios show a weak negative correlation.The overall technology convergence intensity is at a moderate level(averaging 4.51 themes/policy)and displays a new trend of shifting towards lower values.These re-sults provide important data support and a decision-making basis for precisely identifying technological development shortcomings,optimizing industrial chain synergy pathways,and formulating differentiated and forward-looking industrial policies.关键词
氢能政策/技术融合/LDA主题模型Key words
Hydrogen Energy Policy/Technology Convergence/LDA Topic Model分类
社会科学引用本文复制引用
刘璞,胡启萌,董媛..基于LDA主题模型的我国氢能技术融合发展趋势分析[J].科技广场,2025,(5):16-29,14.基金项目
陕西省创新能力支撑计划-科技创新服务体系计划"科技奖励保障及科技安全情报研究能力提升"(项目编号:2024ZG-CXFW-07) （项目编号:2024ZG-CXFW-07）
陕西省科学技术情报学会培育类课题"基于主题识别模型的新兴产业情报研究"(项目编号:2024KTP-25) （项目编号:2024KTP-25）
