year 13, Issue 1 (5-2026)
CJS 2026, 13(1): 80-93 |
Back to browse issues page
Download citation:
BibTeX | RIS | EndNote | Medlars | ProCite | Reference Manager | RefWorks
Send citation to:
BibTeX | RIS | EndNote | Medlars | ProCite | Reference Manager | RefWorks
Send citation to:
Haddad Araghi S, Ghiasi M, Razavi S A A. Mapping the Conceptual Structure of Artificial Intelligence Applications in Libraries: A Scientometric Study Based on Web of Science Data. CJS 2026; 13 (1) :80-93
URL: http://cjs.mubabol.ac.ir/article-1-412-en.html
URL: http://cjs.mubabol.ac.ir/article-1-412-en.html
Department of Knowledge and Information Science, Bab.C, Islamic Azad University, Babol, Iran , Mitra.ghiasi@iau.ac.ir
Abstract: (13 Views)
Background and aim: As artificial intelligence applications have expanded, numerous studies have also investigated the use of this technology in libraries and information centers. This study aims to identify the conceptual structure of artificial intelligence applications in libraries and to analyze the research trends in this field.
Materials and methods: This scientometric study was conducted using co-word analysis. The research dataset consisted of 8,403 documents indexed in the Web of Science database from 2000 to 2024. A total of 30,851 author keywords were extracted, and after cleaning and standardization, 15,360 unique keywords remained. Data were analyzed using BibExcel and VOSviewer. BibExcel was used to prepare the data and construct the keyword co-occurrence matrix, while VOSviewer was used to visualize the co-word network, display the relationship between concepts, and identify thematic clusters based on the VOS clustering technique. Centrality and density indices were additionally utilized to analyze the strategic diagram.
Findings: The findings revealed that among the extracted keywords, the concepts “artificial intelligence” with an occurrence frequency of 1,192, “machine learning” with an occurrence frequency of 969, and “natural language processing” with an occurrence frequency of 643 were the most frequent keywords in the dataset. The strongest co-occurrence relationships between all keyword pairs were observed between "artificial intelligence" and "machine learning," which appeared together 145 times, followed closely by "machine learning" and "natural language processing," with a co-occurrence frequency of 131. Cluster analysis identified eight conceptual clusters. The mean centrality and density values of the clusters were 35.1 and 0.864, respectively. The clusters of “innovative services and user interaction” (78.15; 0.444), “AI ethics and governance” (42.12; 0.679), and “machine learning models and recommender systems” (45.14; 0.836) showed centrality values above the mean. In contrast, the clusters of “large language models and information retrieval” (29.81; 1.59) and “sentiment analysis and misinformation detection” (32.5; 2.32) showed the highest density values. None of the identified clusters were located within the central and well-established zone of the strategic diagram.
Conclusion: Considering the high centrality of clusters pertaining to innovative services, AI ethics and governance, and machine learning models, alongside the high density of clusters focusing on large language models and sentiment analysis, three key requirements emerge for the advancement of this field in libraries: strengthening technological infrastructure, enhancing algorithmic literacy among library professionals, and formulating ethical and operational policies to govern the responsible use of artificial intelligence.
Materials and methods: This scientometric study was conducted using co-word analysis. The research dataset consisted of 8,403 documents indexed in the Web of Science database from 2000 to 2024. A total of 30,851 author keywords were extracted, and after cleaning and standardization, 15,360 unique keywords remained. Data were analyzed using BibExcel and VOSviewer. BibExcel was used to prepare the data and construct the keyword co-occurrence matrix, while VOSviewer was used to visualize the co-word network, display the relationship between concepts, and identify thematic clusters based on the VOS clustering technique. Centrality and density indices were additionally utilized to analyze the strategic diagram.
Findings: The findings revealed that among the extracted keywords, the concepts “artificial intelligence” with an occurrence frequency of 1,192, “machine learning” with an occurrence frequency of 969, and “natural language processing” with an occurrence frequency of 643 were the most frequent keywords in the dataset. The strongest co-occurrence relationships between all keyword pairs were observed between "artificial intelligence" and "machine learning," which appeared together 145 times, followed closely by "machine learning" and "natural language processing," with a co-occurrence frequency of 131. Cluster analysis identified eight conceptual clusters. The mean centrality and density values of the clusters were 35.1 and 0.864, respectively. The clusters of “innovative services and user interaction” (78.15; 0.444), “AI ethics and governance” (42.12; 0.679), and “machine learning models and recommender systems” (45.14; 0.836) showed centrality values above the mean. In contrast, the clusters of “large language models and information retrieval” (29.81; 1.59) and “sentiment analysis and misinformation detection” (32.5; 2.32) showed the highest density values. None of the identified clusters were located within the central and well-established zone of the strategic diagram.
Conclusion: Considering the high centrality of clusters pertaining to innovative services, AI ethics and governance, and machine learning models, alongside the high density of clusters focusing on large language models and sentiment analysis, three key requirements emerge for the advancement of this field in libraries: strengthening technological infrastructure, enhancing algorithmic literacy among library professionals, and formulating ethical and operational policies to govern the responsible use of artificial intelligence.
Keywords: Artificial intelligence, Libraries, Scientometric study, Co-word analysis, Science mapping
Type of Study: Orginal |
Subject:
Scientometrics
Received: 2025/11/24 | Revised: 2026/06/3 | Accepted: 2026/06/8 | ePublished: 2026/06/22
Received: 2025/11/24 | Revised: 2026/06/3 | Accepted: 2026/06/8 | ePublished: 2026/06/22
Send email to the article author
| Rights and permissions | |
 |
This work is licensed under a Creative Commons Attribution-NonCommercial 4.0 International License. |
