| 许欣.基于 BERTopic 和 ARIMA 的文献热点主题及演化趋势分析 —— 以 ESI 材料科学为例[J].情报工程,2026,12(4):093-107 |
| 基于 BERTopic 和 ARIMA 的文献热点主题及演化趋势分析 —— 以 ESI 材料科学为例 |
| Analysis of Hot Topics and Evolution Trends of Literature Based on BERTopic and ARIMA: A Case Study of ESI Materials Science Field |
| |
| DOI： |
| 中文关键词: BERTopic 模型；主题识别；ARIMA；时间序列模型 |
| 英文关键词: BERTopic Model; Topic Identification; ARIMA; Time Series Model |
| 基金项目: |
| 作者 | 单位 | | 许欣 | 江苏科技大学苏州理工学院图书馆 苏州 215600 |
|
| 摘要点击次数: 9 |
| 全文下载次数: 7 |
| 中文摘要: |
| [目的/意义]本研究基于深度学习的主题建模方法，分析 ESI 材料科学领域的热点主题并预测其演化趋势，为科研人员把握学科动态，为决策者优化资源配置提供参考。[方法/过程]利用 BERTopic 模型对 2014―2023 年ESI 材料科学领域的 12138 篇高被引文献进行主题识别，并用 ARIMA 模型对各主题的文献数量进行时间序列预测分析。[结果/结论]研究结果表明，近十年 ESI 材料科学领域的热门主题主要集中在电化学储能、可再生能源的催化反应、柔性可穿戴智能传感器系统、生物材料的医学应用、太阳能电池等方面，充分体现出理工融合、医工融合的特点。在当下及未来一段时间内，ESI 材料科学领域的新能源转化与存储、医学器材等研究主题都将受到广泛关注。 |
| 英文摘要: |
| [Purpose/Significance] This paper analyzes the hot topics in the ESI Materials Science field and predicts their evolution trends based on deep learning topic modeling methods, providing references for researchers to grasp disciplinary
dynamics and decision-makers to optimize resource allocation. [Method/Process] The BERTopic model was used to identify topics from 12138 highly cited papers in the ESI Materials Science field from 2014 to 2023, and the ARIMA model was
used for time-series forecasting analysis of the number of papers for each topic. [Result/Conclusion] The research results show that the hot topics in the ESI Materials Science field over the past decade mainly focus on electrochemical energy storage, catalytic reactions for renewable energy, flexible wearable intelligent sensor systems, medical applications of biomaterials, solar cells, etc., fully reflecting the characteristics of Interdisciplinary subject. In the present and near future, research in the ESI Materials Science field related to new energy conversion and storage and medical devices will continue to receive extensive attention. |
|
查看全文
查看/发表评论 下载PDF阅读器 |
| 关闭 |
