[Objective/Significance] The co-citation network information between journals can reflect their associative relationships. Analyzing this co-citation information for journal clustering aids in achieving a better analysis of disciplinary structure trends, providing reference value for future research by scientific researchers. [Method/Process] This paper proposes a “decoupling-dimensionality reduction” algorithm for journal clustering based on the co-citation matrix. Specifically, it decouples the sample co-citation matrix by incorporating out-of-sample co-citation information, and then utilizes t-SNE and hierarchical clustering methods to reduce the dimensionality and cluster the decoupled co-citation matrix. This paper conducts an empirical analysis on important English journals in the field of management science, collecting co-citation information from the WoS Core Collection database for 905 journals spanning the five-year period from 2018 to 2022 as initial data. [Results/Conclusions] Experimental results indicate that the “decoupling-dimensionality reduction”algorithm for journal clustering based on the co-citation matrix effectively enhances the clustering results compared to traditional co-citation matrix clustering methods, and proves that additional cited network information can provide certain assistance for journal clustering.
Sun Weiyi, Wu Dengsheng.
Research on Journal Clustering Algorithm Based on Decoupling-Dimensionality Reduction of Journal Co-Citation Matrix[J].
Science Focus, 2026, 21(3): 76-89 DOI:10.15978/j.cnki.1673-5668.20251014
本文收集了FMS期刊(Federation of Management Societies of China)英文期刊列表中被Web of Science核心合集(WoS核心集)完整收录的2018—2022年共905种期刊为样本,共计3665399条被引数据,46994种被引来源期刊,样本期刊数据部分统计信息如表2所示,部分数据如表3所示。FMS期刊列表涵盖了管理科学领域的各方面期刊,同时管理科学领域相关研究存在着大量细分的不同方向研究,适合观察期刊聚类结果。
在905种期刊中,总发文量最多的为Journal of Cleaner Production,共23498篇;最少的为International Journal of Entrepreneurial Behavior & Research,共19篇。不同期刊的发文量与被引量存在巨大差距,一部分与期刊的规模发行方式有关,另一部分与期刊的内容与风格有关。
为进一步体现算法修正后聚类标签的合理性,以算法聚类结果中26类为例,表5展示了期刊共被引矩阵“解耦-降维”的期刊聚类算法聚类标签与FMS列表提供的类别。可以看出,26类中的大部分期刊在FMS提供的分类标准中为信息管理,而FMS分类不为信息管理的4种期刊British Journal of Educational Technology, Decision Sciences, Journal of Management in Engineering, Online Information Review都有一定的学科交叉内容,由此可知,“解耦-降维”的期刊聚类算法能够一定程度上识别交叉学科期刊并且提供期刊分类结果的修正参考。
(YangN, HuangF H, ZhangZ Q. Research on the node influence and dissemination in scientific dataset co-citation network[J]. Information Studies: Theory & Application, 2023, 46(11): 45-51.)
[3]
HeD N, WangW J, LuoX G, et al. Cluster analysis of hotspots and research trends of epirubicin-induced cardiotoxicity: a bibliometric study[J]. Frontiers in Pharmacology, 2025, 16: 1616162.
[4]
LeydesdorffL, de Moya-AnegónF, Guerrero-BoteV P. Journal maps, interactive overlays, and the measurement of interdisciplinarity on the basis of Scopus data (1996—2012)[J]. Journal of the Association for Information Science and Technology, 2015, 66(5): 1001-1016.
[5]
ShenL N, XiongB, LiW, et al. Visualizing collaboration characteristics and topic burst on international mobile health research: bibliometric analysis[J]. JMIR Mhealth & Uhealth, 2018, 6(6):e135.
[6]
KocakM, García-ZoritaC, Marugán-LázaroS, et al. Mapping and clustering analysis on neuroscience literature in Turkey: a bibliometric analysis from 2000 to 2017[J]. Scientometrics, 2019, 121(3): 1339-1366.
(WangJ, LiW L. Data clustering research of ESI bibliography from the perspective of journal[J]. Journal of Library and Information Science, 2020, 5(11): 60-65.)
(MengY, WuD S, LiJ P. Model construction and empirical analysis of institute hierarchical clustering based on discipline layout[J]. Journal of Systems Science and Mathematical Sciences, 2019, 39(7): 1054-1067.)
(ChenY Y, XuC M, XuT M. Mechanism of knowledge fusion and value transmission in structural unit of citation-network knowledge fusion[J]. Journal of the China Society for Scientific and Technical Information, 2024, 43(10): 1182-1198.)
[13]
BatageljV, FerligojA, DoreianP. Advances in network clustering and blockmodeling[M]. Hoboken: John Wiley & Sons, Ltd., 2020.
(GouC H, XuH Z. Evolution analysis of investor sentiment research: based on CiteSpace literature co-citation analysis[J]. Management and Administration, 2024(8): 12-20.)
(GaoN, ZhouQ S. Research fronts identification and evolution trend analysis of information science based on co-citation method[J]. Journal of Modern Information, 2024, 44(5): 3-19.)
[18]
KleminskiR, KazienkoP, KajdanowiczT. Analysis of direct citation, co-citation and bibliographic coupling in scientific topic identification[J]. Journal of Information Science, 2022, 48(3): 349-373.
[19]
ŠubeljL, van EckN J, WaltmanL. Clustering scientific publications based on citation relations: a systematic comparison of different methods[J]. PLoS One, 2016, 11(4): e0154404.
[20]
BacciniA, BarabesiL. Interlockingeditorship. A network analysis of the links between economic journals[J]. Scientometrics, 2010, 82(2): 365-389.
[21]
LeydesdorffL. Can scientific journals be classified in terms of aggregated journal-journal citation relations using the Journal Citation Reports[J]. Journal of the American Society for Information Science and Technology, 2006, 57(5): 601-613.
(WenF F, YangQ Q, LiX Y. Research on the interdisciplinary measurement and evolution of Chinese social science and humanity: based on the keywords coupling analysis of the national social science fund projects[J]. Journal of Modern Information, 2022, 42(3): 157-167.)
[24]
ChenG, HongS Q, DuC X, et al. Comparing semantic representation methods for keyword analysis in bibliometric research[J]. Journal of Informetrics, 2024, 18(3): 101529.
[25]
LozanoS, Calzada-InfanteL, Adenso-DíazB, et al. Complex network analysis of keywords co-occurrence in the recent efficiency analysis literature[J]. Scientometrics, 2019, 120(2): 609-629.
[26]
RadhakrishnanS, ErbisS, IsaacsJ A, et al. Novel keyword co-occurrence network-based methods to foster systematic reviews of scientific literature[J]. PLoS One, 2017, 12(3): e0172778.
[27]
ShenL N, XiongB, HuJ M. Researchstatus, hotspots and trends for information behavior in China using bibliometric and co-word analysis[J]. Journal of Documentation, 2017, 73(4): 618-633.
[28]
WangM Y, ChaiL H. Three new bibliometric indicators/approaches derived from keyword analysis[J]. Scientometrics, 2018, 116(2): 721-750.
(YuY, LuoY C, ShiJ, et al. Research on the influence of editorial board on journal research topics: taking information science field as an example[J]. Modern Information, 2024, 44(4): 142-153.)
[31]
LiuA N, LuY W, GongC, et al. Bibliometric analysis of research themes and trends of the co-occurrence of autism and ADHD[J]. Neuropsychiatric Disease and Treatment, 2023, 19: 985-1002.
[32]
NwagwuW E. Bibliographic coupling networks of global research on data literacy by documents, sources and authors[J]. Journal of Librarianship and Information Science, 2025, 57(3): 876-893.
(BuY, WangB L, XuY. Towards the improvements of author bibliographic coupling analysis combined with time information[J]. Journal of Intelligence, 2017, 36(10): 148-151.)
(ZhangR H, YuanJ P. Semantic-and contextual-based author bibliographic coupling analysis[J]. Journal of the China Society for Scientific and Technical Information, 2022, 41(8): 796-811.)
[37]
BuY, NiS K, HuangW B. Combining multiple scholarly relationships with author cocitation analysis: a preliminary exploration on improving knowledge domain mappings[J]. Journal of Informetrics, 2017, 11(3): 810-822.
[38]
MaltsevaD, BatageljV. Citation and bibliographic coupling between authors in the field of social network analysis[J]. Journal of Data and Information Science, 2024, 9(4): 110-154.
[39]
QiuJ P, DongK, YuH Q. Comparative study on structure and correlation among author co-occurrence networks in bibliometrics[J]. Scientometrics, 2014, 101(2): 1345-1360.
[40]
SongY H, WuL J, MaF. A study of differences between all-author bibliographic coupling analysis and all-author co-citation analysis in detecting the intellectual structure of a discipline[J]. The Journal of Academic Librarianship, 2021, 47(3): 102351.
[41]
SongY H, LeiL X, WuL J, et al. Studying domain structure: a comparative analysis of bibliographic coupling analysis and co-citation analysis considering all authors[J]. Online Information Review, 2023, 47(1): 123-137.
(XuY J, LiJ P, WuD S. Identifying the discipline structure with topic coupling analysis of journal article[J]. Library and Information Service, 2024, 68(21): 3-15.)
(MaJ B, XuC. Technology development trend analysis based on paper semantic and citation network[J]. Software Guide, 2024, 23(11): 100-106.)
[46]
GlenissonP, GlänzelW, JanssensF, et al. Combining full text and bibliometric information in mapping scientific disciplines[J]. Information Processing & Management, 2005, 41(6): 1548-1572.
[47]
AhlgrenP, JarnevingB. Bibliographic coupling, common abstract stems and clustering: a comparison of two document-document similarity approaches in the context of science mapping[J]. Scientometrics, 2008, 76(2): 273-290.
(LuC, DongK. Study on adding weight to bibliographic network using citation content: a method based on citation mention[J]. Journal of Intelligence, 2022, 41(11): 171-178.)
(LuW H. Path identification and disruptive innovation of technology R & D based on analysis of patent citation network[J]. Journal of the China Society for Scientific and Technical Information, 2024, 43(9): 1059-1069.)
[53]
孙海生. 基于Web of Science数据库的文献耦合网络实证研究[J]. 情报杂志, 2018, 37(10): 201-207.
[54]
(SunH S. Empirical research on bibliographic coupling networks: based on web of science database[J]. Journal of Intelligence, 2018, 37(10): 201-207.)
[55]
ZhangR H, YuanJ P. Improving bibliographic coupling and co-citation: a study evaluating the effects of various full-text citation features[J/OL]. Journal of Information Science,2025,[2025-10-14]. https://sage.cnpereading.com/doi/10.1177/01655515251322483
[56]
宋歌. 共被引分析方法迭代创新路径研究[J]. 情报学报, 2020, 39(1): 12-24.
[57]
(SongG. Research on iterative innovation path of co-citation analysis method[J]. Journal of the China Society for Scientific and Technical Information, 2020, 39(1): 12-24.)
(HuangF, HouH Y, RenP L, et al. Selection of research front detection methods based on bibliographic coupling and co-citation[J]. Journal of Intelligence, 2018, 37(12): 13-19.)
[60]
Van Der MaatenL, HintonG. Visualizing data using t-SNE[J]. Journal of Machine Learning Research, 2008, 9: 2579-2605.
(LuX L, WuD S. Identifying the structure of disciplines through journal author coupling analysis[J]. Journal of the China Society for Scientific and Technical Information, 2020, 39(11): 1154-1161.)
(ZhangY, PanR, FangK N. Subject topic analysis through community detection in coauthorship network: taking international statistical journals as an example[J]. Quarterly Journal of Economics and Management, 2023, 2(2): 219-240.)
(LuW H, TanZ Y. The measurement of scientific knowledge structure based on journal clustering and deep learning[J]. Journal of the China Society for Scientific and Technical Information, 2020, 39(1): 38-46.)
(YangX Y, YangJ L, YeW H. Network characterization of domain knowledge clusters based on heterogeneous information network[J]. Journal of the China Society for Scientific and Technical Information, 2025, 44(9): 1128-1143.)
(LiX X, LiW. Classification of interdisciplinary academic journals based on knowledge element mining: taking information resource management journals as an example[J]. Chinese Journal of Scientific and Technical Periodicals, 2025, 36(3): 367-374.)
[71]
王方媛, 徐慧婷. 基础学科交叉特征及热点主题识别:以Web of Science大语言模型主题论文为例[J]. 文献与数据学报, 2025, 7(3): 49-63.
[72]
(WangF Y, XuH T. Identifi cation of interdisciplinary characteristics and hot topics of fundamental disciplines: take papers of large language models theme in web of science as an example[J]. Journal of Library and Data, 2025, 7(3): 49-63.)
(QinY, SunL X. Interdisciplinary measurement and institutional analysis based on co-occurrence networks: a case study of health informatics[J]. Journal of Academic Library and Information Science, 2025, 43(4): 103-109.)