Journal of Informetrics最新文献_第3页

Corrigendum to “Automated generation of research workflows from academic papers: a full-text mining framework” [Journal of Informetrics, 19 (2025) 101732] “从学术论文中自动生成研究工作流：一个全文挖掘框架”的勘误表[Journal of informmetrics， 19 (2025) 101732]

IF 3.5 2区管理学 Q2 COMPUTER SCIENCE, INTERDISCIPLINARY APPLICATIONS

Journal of Informetrics

Pub Date : 2025-11-01 DOI: 10.1016/j.joi.2025.101735

Heng Zhang , Chengzhi Zhang

引用次数: 0

Mapping science and revealing disciplinary communication modalities via pre-trained graph neural networks 通过预训练的图神经网络映射科学和揭示学科交流模式

IF 3.5 2区管理学 Q2 COMPUTER SCIENCE, INTERDISCIPLINARY APPLICATIONS

Journal of Informetrics

Pub Date : 2025-11-01 DOI: 10.1016/j.joi.2025.101741

Yujie Zhang , Guoxiu He , Zhuoren Jiang

Current studies predominantly highlight the growing intersections among disciplines but lack insights into more nuanced aspects of science communication. This work investigates disciplinary communication through two metrics: interactivity, defined as the product of knowledge absorption and diffusion, capturing the overall breadth of knowledge interaction; and radiation, the ratio of outward diffusion to absorption, reflecting the relative tendency to export knowledge. To achieve this, we encode the disciplinary information of each paper as a continuous vector by pre-trained graph neural networks on extensive academic data. The metrics are derived from the distances computed using the paper vectors. We categorize the disciplines into four quadrants: “exposed,” “absorptive,” “service,” and “hermetic”, based on the two metrics. Our findings indicate that life-related sciences (medicine, neuroscience) are “exposed,” with open characteristics. Formal sciences (mathematics, physics and astronomy) are “hermetic,” with limited interaction breadth and radiation capacity. Chemistry, business and management are “absorptive,” focusing on knowledge absorption with limited dissemination. Engineering and Energy are “service-oriented,” centered on transformation and connecting. Our findings and computational methods could contribute to a better understanding of scientific communication systems.

目前的研究主要强调学科之间日益增长的交叉，但缺乏对科学传播更细微方面的见解。这项工作通过两个指标来调查学科交流：互动性，定义为知识吸收和扩散的产物，捕捉知识互动的整体广度；辐射，向外扩散与吸收的比率，反映了知识输出的相对倾向。为了实现这一点，我们在广泛的学术数据上通过预训练的图神经网络将每篇论文的学科信息编码为连续向量。度量是从使用纸向量计算的距离中导出的。我们将学科分为四个象限：“暴露的”、“吸收的”、“服务的”和“密封的”，基于这两个指标。我们的研究结果表明，与生命相关的科学（医学、神经科学）是“暴露的”，具有开放的特征。形式科学（数学、物理和天文学）是“封闭的”，相互作用的广度和辐射能力有限。化学、商业和管理是“吸收性的”，注重知识的吸收，传播有限。工程和能源是“面向服务”的，以转化和连接为中心。我们的发现和计算方法有助于更好地理解科学传播系统。

{"title":"Mapping science and revealing disciplinary communication modalities via pre-trained graph neural networks","authors":"Yujie Zhang , Guoxiu He , Zhuoren Jiang","doi":"10.1016/j.joi.2025.101741","DOIUrl":"10.1016/j.joi.2025.101741","url":null,"abstract":"<div><div>Current studies predominantly highlight the growing intersections among disciplines but lack insights into more nuanced aspects of science communication. This work investigates disciplinary communication through two metrics: interactivity, defined as the product of knowledge absorption and diffusion, capturing the overall breadth of knowledge interaction; and radiation, the ratio of outward diffusion to absorption, reflecting the relative tendency to export knowledge. To achieve this, we encode the disciplinary information of each paper as a continuous vector by pre-trained graph neural networks on extensive academic data. The metrics are derived from the distances computed using the paper vectors. We categorize the disciplines into four quadrants: “exposed,” “absorptive,” “service,” and “hermetic”, based on the two metrics. Our findings indicate that life-related sciences (medicine, neuroscience) are “exposed,” with open characteristics. Formal sciences (mathematics, physics and astronomy) are “hermetic,” with limited interaction breadth and radiation capacity. Chemistry, business and management are “absorptive,” focusing on knowledge absorption with limited dissemination. Engineering and Energy are “service-oriented,” centered on transformation and connecting. Our findings and computational methods could contribute to a better understanding of scientific communication systems.</div></div>","PeriodicalId":48662,"journal":{"name":"Journal of Informetrics","volume":"19 4","pages":"Article 101741"},"PeriodicalIF":3.5,"publicationDate":"2025-11-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145415757","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"管理学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

A method for evaluating the productivity and research diversity of an individual scientific research paper 一种评估单个科学研究论文的生产力和研究多样性的方法

IF 3.5 2区管理学 Q2 COMPUTER SCIENCE, INTERDISCIPLINARY APPLICATIONS

Journal of Informetrics

Pub Date : 2025-10-22 DOI: 10.1016/j.joi.2025.101739

Avner Peleg

We develop a method for evaluating the productivity and research diversity of an individual scientific research paper. The method is based on the productivity and research diversity indicator (PRDI), which measures the value of the paper based on a combination of the number of descendant papers of the evaluated paper,

n_{d}

, and the number of research directions emerging from it,

n_{r}

. We define the basic PRDI P by

P = \ln (n_{d} + 1) + n_{r}

, and develop a detailed procedure for calculating it. We then describe an informative generalization of the basic PRDI calculation, which is based on perturbing two of the main steps in the basic calculation. The perturbation generates statistics of the basic PRDI value, and the generalized PRDI,

P^{(g)}

, is defined as the mean value of P for the generalized calculation. We demonstrate the method and the insights that it can provide by applying it for three example papers, which started new research directions within the general research area of the evaluated author. We explain how the method and its ingredients can be used as additional in-depth tools in decisions on tenure, promotion, and hiring in academic institutions, in peer-review of scientific papers, and in analysis of the development of scientific concepts in various areas of science and engineering. Furthermore, our study introduces a number of mathematical methods, such as coarse-graining, perturbation-induced statistics, and coarse-grained genealogical charts, which can be useful in data science applications in general.

我们开发了一种评估单个科研论文的生产力和研究多样性的方法。该方法基于生产力和研究多样性指标（PRDI），该指标根据被评估论文的后代论文数量和由此产生的研究方向数量nr的组合来衡量论文的价值。我们定义了基本PRDI P=ln (nd+1)+nr，并制定了详细的计算程序。然后，我们描述了基本PRDI计算的信息推广，这是基于扰动基本计算中的两个主要步骤。扰动产生基本PRDI值的统计量，广义PRDI P(g)定义为P的均值，用于广义计算。我们通过三篇示例论文来展示该方法及其所能提供的见解，这些论文在被评价作者的一般研究领域内开辟了新的研究方向。我们解释了该方法及其成分如何在学术机构的任期、晋升和招聘决策中，在科学论文的同行评审中，以及在科学和工程各个领域的科学概念发展分析中，作为额外的深入工具使用。此外，我们的研究还介绍了一些数学方法，如粗粒度、微扰诱导统计和粗粒度家谱图，这些方法在数据科学应用中非常有用。

{"title":"A method for evaluating the productivity and research diversity of an individual scientific research paper","authors":"Avner Peleg","doi":"10.1016/j.joi.2025.101739","DOIUrl":"10.1016/j.joi.2025.101739","url":null,"abstract":"<div><div>We develop a method for evaluating the productivity and research diversity of an individual scientific research paper. The method is based on the productivity and research diversity indicator (PRDI), which measures the value of the paper based on a combination of the number of descendant papers of the evaluated paper, <span><math><msub><mrow><mi>n</mi></mrow><mrow><mi>d</mi></mrow></msub></math></span>, and the number of research directions emerging from it, <span><math><msub><mrow><mi>n</mi></mrow><mrow><mi>r</mi></mrow></msub></math></span>. We define the basic PRDI <em>P</em> by <span><math><mi>P</mi><mo>=</mo><mi>ln</mi><mo>⁡</mo><mo>(</mo><msub><mrow><mi>n</mi></mrow><mrow><mi>d</mi></mrow></msub><mo>+</mo><mn>1</mn><mo>)</mo><mo>+</mo><msub><mrow><mi>n</mi></mrow><mrow><mi>r</mi></mrow></msub></math></span>, and develop a detailed procedure for calculating it. We then describe an informative generalization of the basic PRDI calculation, which is based on perturbing two of the main steps in the basic calculation. The perturbation generates statistics of the basic PRDI value, and the generalized PRDI, <span><math><msup><mrow><mi>P</mi></mrow><mrow><mo>(</mo><mi>g</mi><mo>)</mo></mrow></msup></math></span>, is defined as the mean value of <em>P</em> for the generalized calculation. We demonstrate the method and the insights that it can provide by applying it for three example papers, which started new research directions within the general research area of the evaluated author. We explain how the method and its ingredients can be used as additional in-depth tools in decisions on tenure, promotion, and hiring in academic institutions, in peer-review of scientific papers, and in analysis of the development of scientific concepts in various areas of science and engineering. Furthermore, our study introduces a number of mathematical methods, such as coarse-graining, perturbation-induced statistics, and coarse-grained genealogical charts, which can be useful in data science applications in general.</div></div>","PeriodicalId":48662,"journal":{"name":"Journal of Informetrics","volume":"19 4","pages":"Article 101739"},"PeriodicalIF":3.5,"publicationDate":"2025-10-22","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145361873","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"管理学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Does reusing scientific datasets reduce the impact of the papers? 重复使用科学数据集会降低论文的影响力吗？

IF 3.5 2区管理学 Q2 COMPUTER SCIENCE, INTERDISCIPLINARY APPLICATIONS

Journal of Informetrics

Pub Date : 2025-10-19 DOI: 10.1016/j.joi.2025.101737

Bo Yang, Hong Jiao, Qingqing Fan, Jiawen Chen, Jiaxue Liu

Data reuse is increasingly advocated as a strategy to enhance research reproducibility, accelerate project progress, and reduce research costs. Although few dispute the principle of data reuse, its effect on citation performance in experiment-based or data-intensive studies remains uncertain. To dispel concerns about the impact of data reuse on research, researchers require clear evidence of its benefits. This study employs informetric analysis, analysis of variance, and multiple linear regression to conduct a large-scale investigation of scientists’ dataset (re)use behavior, providing direct evidence of the citation performance of their research. The results show that: (i) The volume of released data in biomedical and life sciences continues to grow steadily; however, tracking the (re)use of Gene Expression Omnibus datasets over time shows that actual utilization and reuse have not kept pace with; (ii) Papers that declare the reuse of released datasets, especially those reusing their own data (self-reuse), garner more citations, indicating that dataset reuse does not negatively impact citation performance and may even enhance it; (iii) Our co-citation model predicts that, owing to the “sheep flock effect,” data reuse could increase the exposure of reusers’ related works and subsequently enhance the citation performance of their other publications.

数据重用作为一种提高研究可重复性、加快项目进度和降低研究成本的策略越来越受到推崇。虽然很少有人质疑数据重用的原则，但在基于实验或数据密集型的研究中，数据重用对引文性能的影响仍然不确定。为了消除对数据重用对研究的影响的担忧，研究人员需要明确的证据来证明它的好处。本研究采用信息计量分析、方差分析和多元线性回归等方法，对科研人员的数据集（再）使用行为进行了大规模调查，为科研人员的被引绩效提供了直接证据。结果表明：(i)生物医学和生命科学领域公布的数据量继续稳步增长；然而，随着时间的推移，对基因表达Omnibus数据集的（重复）使用的跟踪表明，实际的利用和重用并没有跟上；（ii）声明重用已发布数据集的论文，特别是那些重用自己数据（自我重用）的论文，获得了更多的引用，这表明数据集重用不会对引用性能产生负面影响，甚至可能提高引用性能；（iii）我们的共被引模型预测，由于“羊群效应”，数据重用可以增加重用者相关作品的曝光率，从而提高其其他出版物的被引性能。

{"title":"Does reusing scientific datasets reduce the impact of the papers?","authors":"Bo Yang, Hong Jiao, Qingqing Fan, Jiawen Chen, Jiaxue Liu","doi":"10.1016/j.joi.2025.101737","DOIUrl":"10.1016/j.joi.2025.101737","url":null,"abstract":"<div><div>Data reuse is increasingly advocated as a strategy to enhance research reproducibility, accelerate project progress, and reduce research costs. Although few dispute the principle of data reuse, its effect on citation performance in experiment-based or data-intensive studies remains uncertain. To dispel concerns about the impact of data reuse on research, researchers require clear evidence of its benefits. This study employs informetric analysis, analysis of variance, and multiple linear regression to conduct a large-scale investigation of scientists’ dataset (re)use behavior, providing direct evidence of the citation performance of their research. The results show that: (i) The volume of released data in biomedical and life sciences continues to grow steadily; however, tracking the (re)use of Gene Expression Omnibus datasets over time shows that actual utilization and reuse have not kept pace with; (ii) Papers that declare the reuse of released datasets, especially those reusing their own data (self-reuse), garner more citations, indicating that dataset reuse does not negatively impact citation performance and may even enhance it; (iii) Our co-citation model predicts that, owing to the “sheep flock effect,” data reuse could increase the exposure of reusers’ related works and subsequently enhance the citation performance of their other publications.</div></div>","PeriodicalId":48662,"journal":{"name":"Journal of Informetrics","volume":"19 4","pages":"Article 101737"},"PeriodicalIF":3.5,"publicationDate":"2025-10-19","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145320001","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"管理学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Document types make the difference 文档类型会造成差异

IF 3.5 2区管理学 Q2 COMPUTER SCIENCE, INTERDISCIPLINARY APPLICATIONS

Journal of Informetrics

Pub Date : 2025-10-15 DOI: 10.1016/j.joi.2025.101738

Zhesi Shen , Robin Haunschild , Lutz Bornmann

引用次数: 0

Predicting technological convergence with multi-channel graph neural networks: A case study of CRISPR 用多通道图神经网络预测技术收敛：以CRISPR为例

IF 3.5 2区管理学 Q2 COMPUTER SCIENCE, INTERDISCIPLINARY APPLICATIONS

Journal of Informetrics

Pub Date : 2025-10-13 DOI: 10.1016/j.joi.2025.101736

Fangjie Xi , Yu Wang , Chenchen Li , Ying Huang , Xiaojun Hu

Identifying emerging technological convergence is essential for anticipating future innovation trajectories. Existing approaches typically rely on either International Patent Classification (IPC) co-occurrence networks, which capture general combination frequencies, or association rule networks, which emphasize statistically significant and often higher-order relationships. However, these two structural views are rarely integrated, limiting their effectiveness in representing both the breadth and depth of technological linkages. To address this gap, we propose a Multi-Channel Graph Convolutional Network (MC-GCN) that treats IPC co-occurrence and association rule networks as structurally distinct inputs. While co-occurrence data reflect raw interaction patterns, association rules—derived via data mining—serve as a refined signal that highlights meaningful and potentially multi-IPC convergence patterns. Our model encodes each view through separate GCN channels and fuses their embeddings within a unified representation space. To establish a comprehensive evaluation, we also include topological link prediction baselines such as Common Neighbors, Adamic–Adar, and Preferential Attachment in our comparative analysis. Applied to CRISPR-related patent data, the MC-GCN significantly outperforms single-channel models, achieving an AUC of 0.973 when combined with XGBoost. Furthermore, five predicted IPC combinations were validated in newly granted patents in early 2025, demonstrating the model’s practical utility in forecasting real-world technological convergence.

识别新兴的技术融合对于预测未来的创新轨迹至关重要。现有方法通常依赖于捕获一般组合频率的国际专利分类（IPC）共现网络，或强调统计显著且通常是高阶关系的关联规则网络。然而，这两种结构观点很少结合起来，限制了它们在表示技术联系的广度和深度方面的有效性。为了解决这一差距，我们提出了一种多通道图卷积网络（MC-GCN），它将IPC共发生和关联规则网络视为结构上不同的输入。虽然共现数据反映了原始的交互模式，但通过数据挖掘派生的关联规则作为精细的信号，突出了有意义的和潜在的多ipc融合模式。我们的模型通过单独的GCN通道对每个视图进行编码，并将它们的嵌入融合到统一的表示空间中。为了建立全面的评价，我们还在比较分析中加入了共同邻居、亚当-阿达尔和优先连接等拓扑链路预测基线。应用于crispr相关专利数据时，MC-GCN显著优于单通道模型，与XGBoost联合使用时AUC达到0.973。此外，2025年初，五种预测的IPC组合在新授予的专利中得到了验证，证明了该模型在预测现实世界技术融合方面的实际效用。

{"title":"Predicting technological convergence with multi-channel graph neural networks: A case study of CRISPR","authors":"Fangjie Xi , Yu Wang , Chenchen Li , Ying Huang , Xiaojun Hu","doi":"10.1016/j.joi.2025.101736","DOIUrl":"10.1016/j.joi.2025.101736","url":null,"abstract":"<div><div>Identifying emerging technological convergence is essential for anticipating future innovation trajectories. Existing approaches typically rely on either International Patent Classification (IPC) co-occurrence networks, which capture general combination frequencies, or association rule networks, which emphasize statistically significant and often higher-order relationships. However, these two structural views are rarely integrated, limiting their effectiveness in representing both the breadth and depth of technological linkages. To address this gap, we propose a Multi-Channel Graph Convolutional Network (MC-GCN) that treats IPC co-occurrence and association rule networks as structurally distinct inputs. While co-occurrence data reflect raw interaction patterns, association rules—derived via data mining—serve as a refined signal that highlights meaningful and potentially multi-IPC convergence patterns. Our model encodes each view through separate GCN channels and fuses their embeddings within a unified representation space. To establish a comprehensive evaluation, we also include topological link prediction baselines such as Common Neighbors, Adamic–Adar, and Preferential Attachment in our comparative analysis. Applied to CRISPR-related patent data, the MC-GCN significantly outperforms single-channel models, achieving an AUC of 0.973 when combined with XGBoost. Furthermore, five predicted IPC combinations were validated in newly granted patents in early 2025, demonstrating the model’s practical utility in forecasting real-world technological convergence.</div></div>","PeriodicalId":48662,"journal":{"name":"Journal of Informetrics","volume":"19 4","pages":"Article 101736"},"PeriodicalIF":3.5,"publicationDate":"2025-10-13","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145319999","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"管理学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Unfinished grants, unending progress: The impact of unfinished research grants on scientific innovation 未完成的资助，永无止境的进步：未完成的研究资助对科学创新的影响

IF 3.5 2区管理学 Q2 COMPUTER SCIENCE, INTERDISCIPLINARY APPLICATIONS

Journal of Informetrics

Pub Date : 2025-10-08 DOI: 10.1016/j.joi.2025.101734

Jiangyang Fu , Xin Liu , Chenwei Zhang , Jiang Li

Scientists may not fulfill the objectives delineated within their research proposals subsequent to the receipt of funding. The extent to which unfinished grants enhance scientific knowledge remains an open question. Drawing upon a dataset from the Research Grants Council of Hong Kong (RGC) that encompasses the years 2010 to 2020, and is distinguished by its inclusion of self-reported grant completion rates, this study seeks to assess the potential contributions of research grants that were not fully completed to the progress of scientific knowledge. The analysis is conducted by leveraging the RGC's detailed records of project completion rates. The results indicate that, notwithstanding a relative lack in productivity and impact, there is no evidence that unfinished grants generate knowledge that is less disruptive than that produced by completed grants. Consequently, it is suggested that funding bodies should consider revising their assessment criteria to recognize the intrinsic merit of grants that are traditionally labeled as unfinished, thus providing more flexibility for the exploration of novel research domains within the grant allocation process.

在收到资助后，科学家可能无法完成其研究计划中所描述的目标。未完成的拨款能在多大程度上增进科学知识，这仍是一个悬而未决的问题。根据香港研究资助局（研资局）2010年至2020年的数据集，本研究旨在评估尚未完全完成的研究资助对科学知识进步的潜在贡献，其特点是纳入了自我报告的资助完成率。这项分析是根据研资局有关项目完成率的详细记录进行的。结果表明，尽管生产力和影响相对缺乏，但没有证据表明未完成的资助产生的知识比完成的资助产生的知识更具破坏性。因此，建议资助机构应考虑修改其评估标准，以认识到传统上被标记为未完成的资助的内在价值，从而在资助分配过程中为探索新的研究领域提供更大的灵活性。

{"title":"Unfinished grants, unending progress: The impact of unfinished research grants on scientific innovation","authors":"Jiangyang Fu , Xin Liu , Chenwei Zhang , Jiang Li","doi":"10.1016/j.joi.2025.101734","DOIUrl":"10.1016/j.joi.2025.101734","url":null,"abstract":"<div><div>Scientists may not fulfill the objectives delineated within their research proposals subsequent to the receipt of funding. The extent to which unfinished grants enhance scientific knowledge remains an open question. Drawing upon a dataset from the Research Grants Council of Hong Kong (RGC) that encompasses the years 2010 to 2020, and is distinguished by its inclusion of self-reported grant completion rates, this study seeks to assess the potential contributions of research grants that were not fully completed to the progress of scientific knowledge. The analysis is conducted by leveraging the RGC's detailed records of project completion rates. The results indicate that, notwithstanding a relative lack in productivity and impact, there is no evidence that unfinished grants generate knowledge that is less disruptive than that produced by completed grants. Consequently, it is suggested that funding bodies should consider revising their assessment criteria to recognize the intrinsic merit of grants that are traditionally labeled as unfinished, thus providing more flexibility for the exploration of novel research domains within the grant allocation process.</div></div>","PeriodicalId":48662,"journal":{"name":"Journal of Informetrics","volume":"19 4","pages":"Article 101734"},"PeriodicalIF":3.5,"publicationDate":"2025-10-08","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145265439","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"管理学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Multidimensional bibliometric assessment of science funding effectiveness 科学资助有效性的多维文献计量评估

IF 3.5 2区管理学 Q2 COMPUTER SCIENCE, INTERDISCIPLINARY APPLICATIONS

Journal of Informetrics

Pub Date : 2025-09-30 DOI: 10.1016/j.joi.2025.101731

Tian-Yuan Huang , Wenjing Xiong

Science funding supports discovery, innovation, and societal progress by enabling research aligned with social needs, but its effectiveness is hard to assess due to the lack of counterfactuals, overlapping funding sources, and varied evaluation metrics. To address this, we developed a multidimensional framework that encompasses research impact, international collaboration, open access status, thematic orientation, and interdisciplinarity, and used it to compare publications funded by the U.S. National Science Foundation (NSF) and the National Natural Science Foundation of China (NSFC) across fifteen natural science fields between 2011 and 2020. Our analysis shows that NSFC support more effectively captures highly cited outputs, whereas NSF funding is more efficient at identifying high impact work and more consistently promotes international partnerships. Both funders have driven substantial growth in open access publishing even though the rising article processing charges threaten equity. In terms of thematic focus, NSFC concentrates on popular research areas while NSF tends to support niche but influential fields. Finally, funded publications consistently demonstrate superior interdisciplinary integration compared to unfunded publications before 2018, indicative of a systemic inclination within financially backed research endeavors to synthesize heterogeneous academic domains for enhanced innovative output. Funded publications outperform unfunded publications both on dimensions of variety and disparity, yet reverses on balance. These findings demonstrate that national funding schemes exert heterogeneous effects on research dynamics, suggesting that future policy should mandate incentives for collaboration and open access, diversify thematic portfolios, and prioritize genuine interdisciplinary innovation.

科学资助通过使研究与社会需求保持一致来支持发现、创新和社会进步，但由于缺乏反事实、重叠的资助来源和不同的评估指标，其有效性难以评估。为了解决这个问题，我们开发了一个多维框架，包括研究影响、国际合作、开放获取状况、专题方向和跨学科性，并使用它来比较2011年至2020年间由美国国家科学基金会（NSF）和中国国家自然科学基金委员会（NSFC）资助的15个自然科学领域的出版物。我们的分析表明，国家自然科学基金的支持更有效地捕获了高被引的产出，而国家自然科学基金的资助更有效地识别了高影响力的工作，并更持续地促进了国际伙伴关系。这两位资助者推动了开放获取出版的大幅增长，尽管不断上涨的文章处理费威胁到了公平。在专题重点方面，国家自然科学基金主要支持热门研究领域，而国家自然科学基金则倾向于支持小众但有影响力的领域。最后，与2018年之前未获得资助的出版物相比，获得资助的出版物始终表现出更好的跨学科整合，这表明在获得资助的研究努力中，系统性倾向于综合异质学术领域，以提高创新产出。资助出版物在多样性和差异方面都优于非资助出版物，但在平衡方面则相反。这些发现表明，国家资助计划对研究动态产生异质效应，表明未来的政策应强制要求鼓励合作和开放获取，使专题组合多样化，并优先考虑真正的跨学科创新。

{"title":"Multidimensional bibliometric assessment of science funding effectiveness","authors":"Tian-Yuan Huang , Wenjing Xiong","doi":"10.1016/j.joi.2025.101731","DOIUrl":"10.1016/j.joi.2025.101731","url":null,"abstract":"<div><div>Science funding supports discovery, innovation, and societal progress by enabling research aligned with social needs, but its effectiveness is hard to assess due to the lack of counterfactuals, overlapping funding sources, and varied evaluation metrics. To address this, we developed a multidimensional framework that encompasses research impact, international collaboration, open access status, thematic orientation, and interdisciplinarity, and used it to compare publications funded by the U.S. National Science Foundation (NSF) and the National Natural Science Foundation of China (NSFC) across fifteen natural science fields between 2011 and 2020. Our analysis shows that NSFC support more effectively captures highly cited outputs, whereas NSF funding is more efficient at identifying high impact work and more consistently promotes international partnerships. Both funders have driven substantial growth in open access publishing even though the rising article processing charges threaten equity. In terms of thematic focus, NSFC concentrates on popular research areas while NSF tends to support niche but influential fields. Finally, funded publications consistently demonstrate superior interdisciplinary integration compared to unfunded publications before 2018, indicative of a systemic inclination within financially backed research endeavors to synthesize heterogeneous academic domains for enhanced innovative output. Funded publications outperform unfunded publications both on dimensions of variety and disparity, yet reverses on balance. These findings demonstrate that national funding schemes exert heterogeneous effects on research dynamics, suggesting that future policy should mandate incentives for collaboration and open access, diversify thematic portfolios, and prioritize genuine interdisciplinary innovation.</div></div>","PeriodicalId":48662,"journal":{"name":"Journal of Informetrics","volume":"19 4","pages":"Article 101731"},"PeriodicalIF":3.5,"publicationDate":"2025-09-30","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145219794","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"管理学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Automated generation of research workflows from academic papers: a full-text mining framework 从学术论文中自动生成研究工作流：一个全文挖掘框架

IF 3.5 2区管理学 Q2 COMPUTER SCIENCE, INTERDISCIPLINARY APPLICATIONS

Journal of Informetrics

Pub Date : 2025-09-22 DOI: 10.1016/j.joi.2025.101732

Heng Zhang , Chengzhi Zhang

The automated generation of research workflows is essential for improving the reproducibility of research and accelerating the paradigm of “AI for Science”. However, existing methods typically extract merely fragmented procedural components and thus fail to capture complete research workflows. To address this gap, we propose an end-to-end framework that generates comprehensive, structured research workflows by mining full-text academic papers. As a case study in the Natural Language Processing (NLP) domain, our paragraph-centric approach first employs Positive-Unlabeled (PU) Learning with SciBERT to identify workflow-descriptive paragraphs, achieving an F₁-score of 0.9772. Subsequently, we utilize Flan-T5 with prompt learning to generate workflow phrases from these paragraphs, yielding ROUGE-1, ROUGE-2, and ROUGE-L scores of 0.4543, 0.2877, and 0.4427, respectively. These phrases are then systematically categorized into data preparation, data processing, and data analysis stages using ChatGPT with few-shot learning, achieving a classification precision of 0.958. By mapping categorized phrases to their document locations in the documents, we finally generate readable visual flowcharts of the entire research workflows. This approach facilitates the analysis of workflows derived from an NLP corpus and reveals key methodological shifts over the past two decades, including the increasing emphasis on data analysis and the transition from feature engineering to ablation studies. Our work offers a validated technical framework for automated workflow generation, along with a novel, process-oriented perspective for the empirical investigation of evolving scientific paradigms. Source code and data are available at: h ttps://github.com/ZH-heng/research_workflow.

研究工作流程的自动化生成对于提高研究的可重复性和加速“科学人工智能”范式至关重要。然而，现有的方法通常只提取碎片化的程序组件，因此无法捕获完整的研究工作流程。为了解决这一差距，我们提出了一个端到端框架，通过挖掘全文学术论文来生成全面、结构化的研究工作流程。作为自然语言处理（NLP）领域的一个案例研究，我们以段落为中心的方法首先使用SciBERT的积极未标记（PU）学习来识别工作流描述段落，获得了0.9772的f1分数。随后，我们利用快速学习的Flan-T5从这些段落中生成工作流短语，分别产生ROUGE-1， ROUGE-2和ROUGE-L得分为0.4543,0.2877和0.4427。然后使用ChatGPT结合few-shot学习将这些短语系统地分为数据准备、数据处理和数据分析三个阶段，分类精度达到0.958。通过将分类短语映射到它们在文档中的文档位置，我们最终生成整个研究工作流程的可读可视化流程图。这种方法有助于分析来自NLP语料库的工作流程，并揭示了过去二十年来主要的方法转变，包括对数据分析的日益重视以及从特征工程到消融研究的转变。我们的工作为自动化工作流生成提供了一个经过验证的技术框架，同时为不断发展的科学范式的实证研究提供了一个新颖的、面向过程的视角。源代码和数据可在：h ttps://github.com/ZH-heng/research_workflow。

{"title":"Automated generation of research workflows from academic papers: a full-text mining framework","authors":"Heng Zhang , Chengzhi Zhang","doi":"10.1016/j.joi.2025.101732","DOIUrl":"10.1016/j.joi.2025.101732","url":null,"abstract":"<div><div>The automated generation of research workflows is essential for improving the reproducibility of research and accelerating the paradigm of “AI for Science”. However, existing methods typically extract merely fragmented procedural components and thus fail to capture complete research workflows. To address this gap, we propose an end-to-end framework that generates comprehensive, structured research workflows by mining full-text academic papers. As a case study in the Natural Language Processing (NLP) domain, our paragraph-centric approach first employs Positive-Unlabeled (PU) Learning with SciBERT to identify workflow-descriptive paragraphs, achieving an F<sub>1</sub>-score of 0.9772. Subsequently, we utilize Flan-T5 with prompt learning to generate workflow phrases from these paragraphs, yielding ROUGE-1, ROUGE-2, and ROUGE-L scores of 0.4543, 0.2877, and 0.4427, respectively. These phrases are then systematically categorized into data preparation, data processing, and data analysis stages using ChatGPT with few-shot learning, achieving a classification precision of 0.958. By mapping categorized phrases to their document locations in the documents, we finally generate readable visual flowcharts of the entire research workflows. This approach facilitates the analysis of workflows derived from an NLP corpus and reveals key methodological shifts over the past two decades, including the increasing emphasis on data analysis and the transition from feature engineering to ablation studies. Our work offers a validated technical framework for automated workflow generation, along with a novel, process-oriented perspective for the empirical investigation of evolving scientific paradigms. Source code and data are available at: h ttps://github.co<em>m/Z</em>H-heng/research_workflow.</div></div>","PeriodicalId":48662,"journal":{"name":"Journal of Informetrics","volume":"19 4","pages":"Article 101732"},"PeriodicalIF":3.5,"publicationDate":"2025-09-22","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145105540","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"管理学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Impact of rhetorical devices on citation behavior: persuasion in scientific papers and its effect on reader response 修辞手段对引文行为的影响：科学论文中的说服及其对读者反应的影响

IF 3.5 2区管理学 Q2 COMPUTER SCIENCE, INTERDISCIPLINARY APPLICATIONS

Journal of Informetrics

Pub Date : 2025-09-18 DOI: 10.1016/j.joi.2025.101729

Kai Meng , Chungwon Koh , Zhejun Zheng , Zhichao Ba , Min Song

Storytelling is an effective method for communicating science, with rhetorical writing offering a structured framework to enrich narratives and enhance their persuasive impact. Scientific writing does not inherently prioritize obscure or convoluted language; however, overly dry and impersonal scientific texts may be difficult to understand and engage with. This study investigates how rhetorical styles in scientific writing influence citation behaviors among readers. Building upon Aristotle's rhetorical theory, we construct computational measures for three key rhetorical strategies in scientific writing: authority (ethos), readability (logos), and emotions (pathos). Using a dataset of over 10 million journal articles from OpenAlex, we analyze the causal relationship between rhetorical styles in scientific writing and their impact on citation behaviors. Our findings reveal that (1) increased use of ethos and pathos in scientific writing positively influences citation counts, while logos has a negative causal effect; (2) author reputation significantly moderates the persuasive effects of rhetoric, particularly mitigating the negative impact of logos; and (3) rhetorical heterogeneity is influenced by factors such as country of publication, publishing formats, disciplines, and citation percentiles. These results offer valuable insights for early-career researchers on effective scientific writing and serve as a reference for publishers developing guidelines.

讲故事是传播科学的有效方法，修辞写作提供了一个结构化的框架来丰富叙述并增强其说服力。科学写作本身不会优先考虑晦涩难懂的语言；然而，过于枯燥和客观的科学文本可能难以理解和参与。本研究旨在探讨科技写作中的修辞风格对读者引用行为的影响。在亚里士多德修辞理论的基础上，我们构建了科学写作中三种关键修辞策略的计算度量：权威（气质），可读性（理性）和情感（感伤）。利用OpenAlex超过1000万篇期刊文章的数据集，我们分析了科学写作中修辞风格及其对引用行为的影响之间的因果关系。研究结果表明：(1)在科学写作中增加使用ethos和pathos对引文数量有积极影响，而logos对引文数量有消极影响；(2)作者声誉显著调节修辞的说服效果，尤其是减缓标志的负面影响；(3)修辞异质性受发表国、出版形式、学科和引文百分位数等因素的影响。这些结果为早期职业研究人员提供了有效科学写作的宝贵见解，并为出版商制定指导方针提供了参考。

{"title":"Impact of rhetorical devices on citation behavior: persuasion in scientific papers and its effect on reader response","authors":"Kai Meng , Chungwon Koh , Zhejun Zheng , Zhichao Ba , Min Song","doi":"10.1016/j.joi.2025.101729","DOIUrl":"10.1016/j.joi.2025.101729","url":null,"abstract":"<div><div>Storytelling is an effective method for communicating science, with rhetorical writing offering a structured framework to enrich narratives and enhance their persuasive impact. Scientific writing does not inherently prioritize obscure or convoluted language; however, overly dry and impersonal scientific texts may be difficult to understand and engage with. This study investigates how rhetorical styles in scientific writing influence citation behaviors among readers. Building upon Aristotle's rhetorical theory, we construct computational measures for three key rhetorical strategies in scientific writing: authority (ethos), readability (logos), and emotions (pathos). Using a dataset of over 10 million journal articles from OpenAlex, we analyze the causal relationship between rhetorical styles in scientific writing and their impact on citation behaviors. Our findings reveal that (1) increased use of ethos and pathos in scientific writing positively influences citation counts, while logos has a negative causal effect; (2) author reputation significantly moderates the persuasive effects of rhetoric, particularly mitigating the negative impact of logos; and (3) rhetorical heterogeneity is influenced by factors such as country of publication, publishing formats, disciplines, and citation percentiles. These results offer valuable insights for early-career researchers on effective scientific writing and serve as a reference for publishers developing guidelines.</div></div>","PeriodicalId":48662,"journal":{"name":"Journal of Informetrics","volume":"19 4","pages":"Article 101729"},"PeriodicalIF":3.5,"publicationDate":"2025-09-18","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145094802","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"管理学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0