情报科学 ›› 2025, Vol. 43 ›› Issue (1): 106-116.

• 业务研究 • 上一篇    下一篇

学术虚拟社区中热点主题识别及其趋势预测方法框架研究

  

  • 出版日期:2025-01-05 发布日期:2025-06-27

  • Online:2025-01-05 Published:2025-06-27

摘要: 【目的/意义】学术虚拟社区已成为当代科研工作者知识开放获取、学术交流和创新的新型重要阵地。基于 学术虚拟社区的热点主题挖掘,推动学科发展动态与领域前沿热点主题的早期探测,有助于学者尽早把握学科发 展历程和预判未来的学科热点方向,进而支撑与服务于科研管理决策。【方法/过程】本研究聚焦学术虚拟社区中的 知识交流,以探测科研工作者的需求为出发点,旨在形成具有通用性、迁移性、关注语义信息的学术虚拟社区中热 点主题识别及其趋势预测的方法框架,并基于生物信息学领域的学术虚拟社区开展实证。【结果/结论】本研究从热 点主题演化的前端(需求端)进行识别与趋势预测,能够有效降低科技文献时滞性对热点预测带来的负面影响,实 证结果表明生物信息学领域论文中的热点主题比社区中的热点主题滞后2年。【创新/局限】在对推动热点主题的前 瞻性预测工作中是一次有益的探索,而在实证研究的应用范围、多源异构数据的融合分析等方面有待完善。

Abstract: 【Purpose/significance】Academic virtual community has become a new important position for contemporary researchers to open access to knowledge, academic exchange and innovation. Based on the mining of hot topics in academic virtual communities, pro⁃ moting the early detection of discipline development trends and frontier hot topics in the field will help scholars grasp the discipline development process and predict the future direction of discipline hot topics as soon as possible, and then support and serve scientific research management decisions.【Method/process】This study focuses on the knowledge exchange in academic virtual communities. Starting from the needs of scientific researchers, it aims to form a method framework for identifying hot topics and predicting their trends in the perspective of academic virtual communities with versatility, mobility and attention to semantic information. Based on the academic virtual community in the field of bioinformatics, empirical research is carried out.【Result/conclusion】This study identifies and predicts the trend from the front end of the evolution of hot topics (demand side), which can effectively reduce the negative impact of the time lag of scientific literature on hot topic prediction The empirical results show that the hot topics in the papers in the field of bioinformatics lag behind the hot topics in the community by 2 years.【Innovation/limitation】It is a useful exploration in the forwardlooking prediction of hot topics, but it needs to be improved in the application scope of empirical research and the fusion analysis of multi-source heterogeneous data.