Scientometrics最新文献

Many claims have been made in the past that Management and Organization Studies (MOS) is becoming increasingly fragmented, and that this fragmentation is causing it to drift into self-reference and irrelevance. Despite the weight of this claim, it has not yet been subjected to a systematic empirical test. This paper addresses this research gap using the tribalization approach and diachronic co-citation analyses. Based on 22,430 papers published in 14 MOS journals between 1980 and 2019, we calculate local and global centrality measures and the flow of cited articles between co-citation communities over time. In addition, we use a node-removal strategy to test whether only ritualized citations ensure MOS cohesion. Rather than tribalization, our results suggest a center–periphery structure. Furthermore, more peripheral papers are integrated into the central co-citation communities, but the lion's share of the flow of cited papers occurs over time to only a small number of large clusters. An increase of fragmentation and crowding-out of smaller clusters in MOS in seen in the polycentrically organized core 2014–2019.

As the volume of scientific literature expands rapidly, accurately gauging and predicting the citation impact of academic papers has become increasingly imperative. Citation counts serve as a widely adopted metric for this purpose. While numerous researchers have explored techniques for projecting papers’ citation counts, a prevalent constraint lies in the utilization of a singular model across all papers within a dataset. This universal approach, suitable for small, homogeneous collections, proves less effective for large, heterogeneous collections spanning various research domains, thereby curtailing the practical utility of these methodologies. In this study, we propose a pioneering methodology that deploys multiple models tailored to distinct research domains and integrates early citation data. Our approach encompasses instance-based learning techniques to categorize papers into different research domains and distinct prediction models trained on early citation counts for papers within each domain. We assessed our methodology using two extensive datasets sourced from DBLP and arXiv. Our experimental findings affirm that the proposed classification methodology is both precise and efficient in classifying papers into research domains. Furthermore, the proposed prediction methodology, harnessing multiple domain-specific models and early citations, surpasses four state-of-the-art baseline methods in most instances, substantially enhancing the accuracy of citation impact predictions for diverse collections of academic papers.

Metaphors play a crucial role in facilitating the comprehension and analysis of knowledge. “Knowledge as energy” is a well-established metaphorical framework that provides unique benefits for comprehending the dissemination of knowledge and enabling its quantification. Nevertheless, empirical studies employing this framework are limited, especially in the area of the work–energy metaphor, which primarily remains theoretical. This paper proposes an application scheme for the work– energy metaphor in interdisciplinary citation analysis. In this scheme, disciplines are considered entities; various factors that drive the progress of a discipline are considered forces; energy is considered the knowledge produced or transferred in the citations. Building upon the work–energy theorem in physics, this study developed indicators reflecting citation quality and velocity to assess interdisciplinary research progression. An empirical investigation was carried out, utilizing these indicators to evaluate the influence of interdisciplinary citations on disciplines. In the experiments, we used Library and Information Science (LIS) from 2012 to 2021 as an example to analyze the impact of interdisciplinary citations from LIS on other disciplines over two time periods. The experiments demonstrated the feasibility of the work–energy metaphorical framework proposed in this paper. It was also found that Computer Science, Management, and Business experienced the highest impact from LIS interdisciplinary citations and exhibited steady growth over a 10-year period. Environmental Science has substantial potential for the future.

This investigation explores the impact of geographical names within article titles on citation frequency across a corpus of literature within the field of library and information science, spanning from 2018 to 2020, and encompassing 56 journal titles. We hypothesized that the presence of geographical names of nations in article titles would negatively correlate with citation counts. Our primary analysis of 1330 articles with geographical names in titles versus 8702 without, revealed a statistically significant, albeit small, difference in median citations, favoring articles without geographical names (mdn = 7) over those with geographical names (mdn = 6). Contrary to our secondary hypothesis, a proximity analysis demonstrated a weak, positive correlation between the position of geographical names near the title end and citation counts. Our examination found little evidence supporting differential citation frequency based on the Human Development Index (HDI) of the nations mentioned in titles. However, although a journal’s impact score strongly predicted citation counts for articles, we found that these counts were depressed when articles in those journals contained a geographic name. We found a negative correlation between the frequency of geographical names in article titles and the journals’ impact scores, yet this was weakly, statistically significant. Our data also suggested a vague positional preference for nations within titles, unrelated to HDI. Furthermore, the likelihood of journals publishing articles mentioning nations of varying HDI was found to be statistically insignificant. This study sheds light on the nuanced influence of title specificity, through geographical names, on scholarly communication and citation impact, indicating a slight preference for broader title phrasing in garnering citations.

Literature retrieval helps scientists find previous work that is relative to their own research or even get new research ideas. However, the discrepancy between retrieval results and the ultimate intention of citation is neglected by most literature retrieval models. Citation intent refers to the researcher’s motivation for citing a paper. A citation intent graph with homogeneous nodes and heterogeneous hyperedges can represent different types of citation intents. By leveraging the citation intent information included in a hypergraph, a retrieval model can guide researchers on where to cite its retrieval result by understanding the citation behaviour in the graph. We present a ranking model called CitenGL (Citation Intent Graph Learning) that aims to extract citation intent information and textual matching signals. The proposed model consists of a heterogeneous hypergraph encoder and a lightweight deep fusion unit for efficiency trade-offs. Compared to traditional literature retrieval, our model fills the gap between retrieval results and citation intention and yields an understandable graph-structured output. We evaluated our model on publicly available full-text paper datasets. Experimental results show that CitenGL outperforms most existing neural ranking models that only consider textual information, which illustrates the effectiveness of integrating citation intent information with textual information. Further ablation analyses show how citation intent information complements text-matching signals and citation networks.

This study critically evaluates gender assignment methods within academic contexts, employing a comparative analysis of diverse techniques, including a SVM classifier, gender-guesser, genderize.io, and a Cultural Consensus Theory based classifier. Emphasizing the significance of transparency, data sources, and methodological considerations, the research introduces nomquamgender, a cultural consensus-based method, and applies it to Teseo, a Spanish dissertation database. The results reveal a substantial reduction in the number of individuals with unknown gender compared to traditional methods relying on INE data. The nuanced differences in gender distribution underscore the importance of methodological choices in gender studies, urging for transparent, comprehensive, and freely accessible methods to enhance the accuracy and reliability of gender assignment in academic research. After reevaluating the problem of gender imbalances in the doctoral system we can conclude that it’s still evident although the trend is clearly set for its reduction. Finaly, specific problems related to some disciplines, including STEM fields and seniority roles are found to be worth of attention in the near future.

Peer review is a critical process used in academia to assess the quality and validity of research articles. Top-tier conferences in the field of artificial intelligence (e.g. ICLR and ACL et al.) require reviewers to provide confidence scores to ensure the reliability of their review reports. However, existing studies on confidence scores have neglected to measure the consistency between the comment text and the confidence score in a more refined way, which may overlook more detailed details (such as aspects) in the text, leading to incomplete understanding of the results and insufficient objective analysis of the results. In this work, we propose assessing the consistency between the textual content of the review reports and the assigned scores at a fine-grained level, including word, sentence and aspect levels. The data used in this paper is derived from the peer review comments of conferences in the fields of deep learning and natural language processing. We employed deep learning models to detect hedge sentences and their corresponding aspects. Furthermore, we conducted statistical analyses of the length of review reports, frequency of hedge word usage, number of hedge sentences, frequency of aspect mentions, and their associated sentiment to assess the consistency between the textual content and confidence scores. Finally, we performed correlation analysis, significance tests and regression analysis on the data to examine the impact of confidence scores on the outcomes of the papers. The results indicate that textual content of the review reports and their confidence scores have high level of consistency at the word, sentence, and aspect levels. The regression results reveal a negative correlation between confidence scores and paper outcomes, indicating that higher confidence scores given by reviewers were associated with paper rejection. This indicates that current overall assessment of the paper’s content and quality by the experts is reliable, making the transparency and fairness of the peer review process convincing. We release our data and associated codes at https://github.com/njust-winchy/confidence_score.

Investigating the intricate relationship between citation similarity and the citation interval offers vital insights for refining citation recommendation systems and enhancing citation evaluation models. This is also a new perspective for understanding citation patterns. In this study, we used the Library and Information Science (LIS) field as an example to determine and discuss the correlation between citation similarity and the citation interval. Using the methods of data collection, paper title preprocessing, text vectorization based on simCSE, calculation of citation similarity and the citation interval, and calculation of the index per citing paper, this study found the following LIS domain-based results: (i) there is a significant negative correlation between citation similarity and the citation interval, but the correlation coefficient is low. (ii) The citation intervals of the least relevant series of cited papers exhibit a more pronounced susceptibility to citation similarity than the most relevant series of cited papers. (iii) The citation intervals of the most relevant cited papers are more concentrated within 12 years and more likely to be published within the average citation interval, typically from the newer half of the cited paper list and published later within 5 years of the citation half-life. This study concludes that researchers usually pay more attention to the latest and most cutting-edge and strongly relevant existing research than to weakly relevant existing research. Continuous attention and timely incorporation of knowledge into the research direction will promote a more rapid and specialized diffusion of knowledge. These findings are influenced by the accelerated dissemination of information via Internet, heightened academic competition, and the concentration of research endeavors in specialized disciplines. This study not only contributes to the scholarly discussion of citation analysis but also lays the foundation for future exploration and understanding of citation patterns.

Altmetrics have led to new quantitative studies of science through social media interactions. However, there are no models of science communication that respond to the multiplicity of non-academic channels. Using the 3653 authors with the highest volume of altmetrics mentions from the main channels (Twitter, News, Facebook, Wikipedia, Blog, Policy documents, and Peer reviews) to their publications (2016-2020), it has been analyzed where the audiences of each discipline are located. The results evidence the generalities and specificities of these new communication models and the differences between areas. These findings are useful for the development of science communication policies and strategies.

Coauthor and acknowledgment data were captured for 1384 research articles published between 1980 and June, 2023 that focused on tardigrades. Articles indexed in Web of Science or an archives of tardigrade literature were downloaded and thoroughly examined for personal acknowledgment data. Annual publication counts and coauthor maps for four successive time periods (1980–1999, 2000–2008, 2009–2017, 2018-June 2023) showed growth in the literature and increased research activity (more researchers, more complex networks, more international collaboration), beginning in 2000. A two-level Personal Acknowledgments Classification (PAC), was used to code types of acknowledgments. The majority of articles focused on field studies and/or descriptions of new species of tardigrades. This was reflected in rankings of acknowledgment categories and additions to the PAC. Ranked lists of frequently-thanked acknowledgees (all tardigrade researchers) were produced for each period. Acknowledgment profiles of four frequently-thanked researchers identified three different roles that researchers might play in tardigrade studies—”informal academic editorial consultant,” “taxonomic gatekeeper,” and “all-rounder.” Acknowledgments honoring people by naming a new species after them were only found in the species description, not in the formal acknowledgment section.