（本文在线发表于《Journal of the Australian Library and Information Association》2025年4月）

ABSTRACT：Traditional research evaluation relies heavily on external indicators such as citations and social media mentions. This study proposes an alternative approach by evaluating the impact of researchers through their research content, introducing internal evaluation indicators. Three author-level topic-based evaluation indicators were established to evaluate researchers’ innovative impact, using the topic model BERTopic, the Jensen-Shannon divergence and interdisciplinary measurement. Based on these indicators, a multidimensional evaluation framework that systematically integrates innovative impact with established academic and social impact dimensions was constructed. An empirical analysis of researchers in the fields of scientometrics and informetrics reveals two key findings. First, researchers’ innovative impact can be assessed by quantifying research topics in terms of novelty, interdisciplinarity and relevance. Second, innovative impact can be used as a complementary evaluation dimension to traditional academic impact metrics. This study enriches research evaluation by introducing internal research content-based indicators, expanding the dimensions of academic impact, and providing new methods and ideas for research evaluation.

摘要：传统科研评价主要依赖引用量、社交媒体提及量等外部指标。本研究提出一种通过研究内容评估学者影响力的替代方案，引入内部评价指标。基于主题模型BERTopic、Jensen-Shannon散度与跨学科测度方法，构建了三个作者层面的主题评价指标以评估学者创新影响力，进而建立系统性融合创新影响力与传统学术/社会影响力的多维评价框架。以科学计量与信息计量领域学者的实证分析揭示两个重要发现：其一，通过研究主题的新颖性、跨学科性和相关性量化，可实现学者创新影响力的有效评估；其二，创新影响力可作为传统学术评价维度的有效补充。本研究通过引入基于研究内容的内部评价指标，拓展了学术影响力的评价维度，为研究评价提供了新方法与新思路。

链接：Indicators for Evaluating the Innovative Impact of Researchers Based on Research Topics: https://www.tandfonline.com/doi/full/10.1080/24750158.2025.2483657?src=exp-la 

转载本文请联系原作者获取授权，同时请注明本文来自刘桂锋科学网博客。

链接地址：https://wap.sciencenet.cn/blog-39723-1481192.html?mobile=1

收藏

分享