节点文献
基于AutoPhrase-SBERT论文-专利文本挖掘的技术演化研究
A Study of Technological Evolution of Paper-patent Text Mining Based on AutoPhrase - SBERT
【摘要】 [目的/意义]通过对制氢领域的论文和专利文本分析,揭示该领域的技术演化趋势,为科研人员、工程师了解该领域的关键技术和发展趋势提供参考。[方法/过程]采用AutoPhrase-SBERT算法对2000—2022年期间我国制氢领域的论文和专利进行主题提取与识别。通过皮尔逊相关系数对相邻时间窗口下论文-专利主题进行关联计算,并利用桑基图对主题关联性进行可视化,从而揭示我国制氢领域技术演变过程。[结果/结论]该方法能够展示领域技术随时间的变化趋势,并能揭示论文和专利主题之间的关系和相互影响。
【Abstract】 [Purpose/significance] Through the analysis of the papers and patent texts within the hydrogen production field, this paper reveals the technological evolution trend, which provides references for researchers and engineers to understand the key technologies and development trends in this field. [Method/process] This paper uses the AutoPhrase-SBERT algorithm to extract and identify the topics of papers and patents in the field of hydrogen production from 2000 to 2022 in China. And it employs Pearson correlation coefficients to calculate the correlation between the paper and patent subject under the adjacent time window, and employs Sankey diagrams to visualize the topic correlation, so as to reveal the technological evolution in the field of hydrogen production in China. [Result/conclusion] This method can show the changing trend of the technology in this field over time, and can reveal the relationship and interaction between the paper and patent topics.
【Key words】 AutoPhrase; SBERT; text mining; technological evolution;
- 【文献出处】 情报探索 ,Information Research , 编辑部邮箱 ,2024年08期
- 【分类号】G255.53;TQ116.2
- 【下载频次】5