تحلیل بروندادهای علمی پایگاه وب آوساینس در حوزه سواد هوش مصنوعی در کتابخانه

Purpose: The digital revolution and the remarkable advancements in artificial intelligence (AI) have brought about profound transformations across all facets of human life. Libraries, as social institutions preserving knowledge and serving as gateways to information, play a pivotal role in these transformations. The expanding applications of artificial intelligence offer unprecedented opportunities to advance libraries' social goals and missions, while also enhancing their services to users. Librarians, as information intermediaries, bear a crucial responsibility in promoting AI literacy and assisting users in the conscious and ethical use of artificial intelligence technologies. There is still no clear outlook on the status of research related to AI literacy in library environments. Given this imperative, the current study analyzes the scholarly output of the Web of Science database concerning AI literacy in libraries to identify the accessibility of resources, predominant research approaches and methodologies, subject areas, and key concepts within this domain.
Methodology: This research was conducted using a descriptive-analytical approach within the field of scientometrics. The study's statistical population comprised 319 scholarly articles indexed in the Web of Science Citation database, spanning the period from 1975 to 2024. To identify relevant data, the Systematic Review methodology was employed, utilizing the methodological framework proposed by Kitchenham and Charters, which was applied in three phases: Planning, Conducting, and Reporting the review. To identify relevant prior studies, three keyword groups were defined and searched in the Web of Science citation database's advanced search. This was achieved using a combined search strategy with Boolean operators, selecting the Title (TI), Abstract (AB), Keywords (AK), Topic(TS), and Research Area (SU) tags. Furthermore, four inclusion criteria were established for selecting articles: 1. The article's focus on the field of AI literacy in libraries and its related terminology; 2. The presence of searched keywords or their equivalents in the article's title, abstract, or keywords; 3. The research-oriented nature of the articles; and 4. Availability of full-text articles. Accordingly, the initial data collection was conducted on July 29, 2024. To evaluate research quality and ensure a precise and transparent report, the PRISMA 2020 checklist was utilized. Subsequently, articles were assessed based on the study's criteria through three phases: identification, screening, and inclusion. Ultimately, 35 articles were deemed eligible for inclusion in the study. Subsequently, the required research data was extracted using a dedicated worksheet. Data analysis was then conducted utilizing Excel and VOSviewer software.
Findings: A study of scholarly outputs in AI literacy within libraries revealed that 62.9% of the articles in this domain were published via hybrid open access. Following this, gold open access articles ranked second (28.6%), with bronze open access articles placing third (5.7%). Conversely, only a small fraction of scholarly articles (2.9%) remain under subscription-based access. The predominant research approach has been quantitative (45.71%). This has primarily been achieved through reliance on survey method (57.14%). The majority of the reviewed scientific articles (80%) fall within the "Information and library Science" subject area. Furthermore, in the co-occurrence analysis of keywords, "academic libraries" and "artificial intelligence" were identified as the most frequently occurring and prominent components within the thematic clusters.
Conclusion: Findings indicate that the field of AI literacy in libraries aligns with the global trend of open access. This facilitates access for diverse stakeholders, including librarians, researchers, students, and policymakers, significantly contributing to the development of this nascent domain. However, the prevalence of hybrid open access models underscores the need to consider the economic and ethical implications of associated funding models. Supplementary strategies, such as green open access (self-archiving) and direct correspondence, play a crucial role in preserving the comprehensiveness of research datasets. The prevalence of a quantitative approach, specifically through survey methods, in research on AI literacy in libraries indicates that researchers have prioritized measuring and describing the current state of this nascent field. However, data analysis reveals a notable balance between quantitative and qualitative approaches. This balanced distribution suggests that researchers have not only sought to describe the existing situation but also to gain a deeper understanding of the experiences and concepts involved. This demonstrates that from the outset, researchers have recognized the multifaceted nature of AI literacy and have strived to examine this emerging phenomenon from various angles, which contributes to the more robust and precise development of the field. Topical analysis reveals a primary focus on Information and library Science, with Computer Science providing a foundational and vital role. Co-occurrence network analysis of keywords reveals that "academic libraries" and "artificial intelligence" have received widespread attention. Nevertheless, "services" functions as the central axis of this network, indicating researchers' focus on developing AI applications to enhance library services. Although this approach holds promise, it is accompanied by challenges such as rapid technological changes, organizational resistance, high implementation costs, a shortage of specialized personnel, and the need for continuous training of users and staff.