Promoting Education for Sustainable Development through AI-Enhanced Library Information Retrieval Systems

Promoting Education for Sustainable Development through AI-Enhanced Library Information Retrieval Systems | Research Square window.SnipcartSettings = { analytics: { enabled: false } }; (function() { var accessVector = localStorage.getItem('access_vector') || ''; window.dataLayer = window.dataLayer || []; if (accessVector) { window.dataLayer.push({ user: { profile: { profileInfo: { snid: accessVector } } } }); } })(); (function(w,d,s,l,i){w[l]=w[l]||[];w[l].push({'gtm.start':new Date().getTime(),event:'gtm.js'});var f=d.getElementsByTagName(s)[0],j=d.createElement(s),dl=l!='dataLayer'?'&l='+l:'';j.async=true;j.src='https://www.googletagmanager.com/gtm.js?id='+i+dl;f.parentNode.insertBefore(j,f);})(window,document,'script','dataLayer','GTM-K279D39R'); Browse Preprints In Review Journals COVID-19 Preprints AJE Video Bytes Research Tools Research Promotion AJE Professional Editing AJE Rubriq About Preprint Platform In Review Editorial Policies Our Team Advisory Board Help Center Sign In Submit a Preprint Cite Share Download PDF Article Promoting Education for Sustainable Development through AI-Enhanced Library Information Retrieval Systems Yunping Zhao, Zhou Lan, Chengfei Wei This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-8751190/v1 This work is licensed under a CC BY 4.0 License Status: Under Review Version 1 posted 15 You are reading this latest preprint version Abstract In an era where the international community places significant emphasis on sustainable development goals, the strategic importance of education for sustainable development (ESD) is particularly pronounced. In the Industry 5.0 era, a growing number of university libraries have adopted Large-Scale Artificial Intelligence Models. This study surveyed 315 students and employed a multiple regression model to investigate the mechanisms through which the application of Large-Scale Artificial Intelligence Models to library information retrieval systems impacts ESD. The findings indicate that the integration of Large-Scale Artificial Intelligence Models into library information retrieval systems can enhance ESD while also improving users' information assessment capability, digital literacy, and awareness of privacy and security. Concurrently, the aforementioned skills, literacy, and awareness of uses served as mediating factors. These results remained consistent after undergoing a series of robustness tests. This study provides technical support and practical paradigms for enhancing the inclusiveness, precision, and long-term impact of education for sustainable development, as well as for cultivating interdisciplinary talents with a global perspective and green development concepts. Social science/Education Business and commerce/Information systems and information technology Social science/Science technology and society Large-Scale Artificial Intelligence Models Library Information Retrieval System Education for Sustainable Development Sustainable Development Industry5.0 1. Introduction Education is a cornerstone of sustainable development, playing an indispensable role in advancing the global transition toward a sustainable future. In recent years, education for sustainable development (ESD) has been increasingly recognized as a vital means of achieving sustainability, with its contributions to quality education and lifelong learning gaining acknowledgment from national societies, major countries and regions, and other stakeholders. ESD was designated as one of the United Nations' Sustainable Development Goals in 2015. In January 2016, the 2030 Agenda for Sustainable Development explicitly called for inclusive and equitable quality education across various stages, including early childhood, primary, secondary, and higher education, as well as technical and vocational training. It urged countries to provide financial and technical support, develop policies and programs for high-quality distance learning in higher education, and encourage early student engagement with science, technology, engineering, and mathematics fields [ 1 ]. Education for Sustainable Development: Towards achieving the SDGs (ESD for 2030) is the theme of the third phase of UNESCO's global action plan for ESD. This plan was approved in September 2019 and officially launched in June 2020. Its accompanying document, Education for Sustainable Development: A Roadmap, highlights key areas, including but not limited to: the intersection of ESD and the sustainable development goals, transformative actions and ESD, structural issues and ESD, and technological advancements and ESD. Within the dimension of technological advancements and ESD, critical areas of exploration include the impact of artificial intelligence, the Fourth Industrial Revolution, and the pathways through which new technologies can empower sustainable education [ 2 ]. These themes provide the broader context for this study. The widespread adoption of Large-Scale Artificial Intelligence Models and other digital technologies is driving the digital transformation of education [ 3 ]. Within this context, libraries, as critical hubs for knowledge dissemination and lifelong learning, play a pivotal role. The iterative enhancement of their information retrieval systems has a profound impact on advancing ESD. This form of education seeks to cultivate versatile individuals equipped with environmental awareness, social responsibility, and innovative capabilities [ 4 ]. As repositories of knowledge, the optimization of libraries' digital services directly influences the accessibility and dissemination efficiency of educational resources for sustainable development. To better meet the increasingly diverse academic needs of students and faculty, there is an urgent need to advance the intelligent upgrading of library information retrieval systems. Currently, AI technologies are experiencing rapid advancement. LLMs, as a groundbreaking achievement in the field of AI, are deeply permeating various industries. Trained on vast datasets, LLMs demonstrate exceptional capabilities in natural language understanding, generation, and knowledge association, enabling precise analysis of complex semantics and the identification of latent connections of different pieces of information. In information retrieval applications, Large-Scale Artificial Intelligence Models overcome the limitations of traditional retrieval systems that rely on keyword matching, offering greater accuracy in interpreting user needs and organizing knowledge resources. An increasing number of libraries are adopting Large-Scale Artificial Intelligence Models. The new generation of library information retrieval systems, leveraging the algorithmic advantages of these models, can rapidly locate vast online academic resources and digitized literature. Additionally, these systems can construct interdisciplinary knowledge graphs, accurately integrating research outcomes and practical case studies from the field of sustainable developments, including spanning environmental science, social governance, economic transformation, and other dimensions, thereby providing learners with a systematic knowledge framework. Furthermore, library information retrieval systems transcend geographical and temporal constraints, enabling educators and students in remote areas to access high-quality educational resources equitably, thus promoting the inclusive sharing of educational resources for sustainable development. Additionally, these systems can incorporate interactive learning modules, facilitating activities such as knowledge quizzes and case study discussions related to sustainable development themes. This enhances user engagement and fosters a widespread commitment to sustainability. Consequently, libraries are transitioning into intelligent knowledge service hubs, providing sustained momentum for achieving the goals of ESD. Currently, research exploring the connection between Large-Scale Artificial Intelligence Models and ESD is in its preliminary stages, yet significant gaps remain [ 5 – 6 ]. In particular, the impact of applying Large-Scale Artificial Intelligence Models in specific contexts on ESD remains underexplored. Existing research primarily focuses on the superficial impact of technological applications on knowledge acquisition models, with limited exploration of the long-term mechanisms through which the transformation of library information retrieval ecosystems, driven by Large-Scale Artificial Intelligence Models, shapes ESD [ 7 ]. Furthermore, when integrating AI literacy frameworks, there is a lack of detailed analysis regarding the differential impact pathways for users of varying genders, regions, and educational attainment levels. It remains unclear how the application of Large-Scale Artificial Intelligence Models in library information retrieval systems can precisely address the diverse needs of ESD. Addressing this research gap is critical for the decision-making and implementation of digital library initiatives. This study examines 'the impact of applying Large-Scale Artificial Intelligence Models to library information retrieval systems on ESD'. Through empirical methods, it analyzes the effects, pathways, and heterogeneity of these impacts across groups differentiated by gender, region, and levels of AI education. The research integrates the Technology Acceptance Model, AI literacy framework, and sustainable development theory to explore the relationships among users' perceptions, applications, trust in AI, and ESD, thereby enhancing explanatory power. Path analysis is employed to identify and quantify the influence of each factor on ESD. This study offers practical recommendations for the digital transformation of university libraries, providing actionable guidance for optimizing digital services and leveraging Large-Scale Artificial Intelligence Models to deepen ESD practices and enhance educational empowerment. The structure of this study is organized as follows: following a literature review, hypotheses are proposed based on the Technology Acceptance Model, AI literacy framework, and sustainable development theory. Subsequently, the study elaborates on sample selection, data sources, variable definitions, and model construction. The regression results of the model are then analyzed, followed by robustness and heterogeneity tests of the applied models. Finally, the study concludes with a summary of findings and offers policy recommendations. 2. Related work 2.1. Utilization of Large-Scale Artificial Intelligence Models in Library Information Retrieval Systems Currently, the rapid advancement of Large-Scale Models is driving a new wave of innovation in artificial intelligence. To advance the digital transformation of libraries, the Chinese government has introduced a series of policies to strengthen the strategic development of smart libraries. Within the framework of 'Large-Scale Models + Smart Libraries', Large-Scale Artificial Intelligence Models serve as the core for extending smart library service applications. These services provide readers with a unified access point, offering diverse functionalities such as consultation and question-answering, search recommendations, and innovative reading experiences [ 8 ]. As key hubs for knowledge innovation and talent cultivation, university libraries play a critical role in supporting academic research, serving the learning needs of students and faculty, and advancing education. The deployment of Large-Scale Artificial Intelligence Models in information retrieval systems has become an inevitable choice for these libraries to align with contemporary technological trends. Currently, university libraries are actively implementing localized Large-Scale Artificial Intelligence Models as a strategic measure to deepen digital transformation and innovate service models. According to the 2024 annual report of the International Federation of Library Associations and Institutions (IFLA), over 78% of libraries worldwide have initiated digital transformation programs, with Chinese public libraries leading globally, achieving a digital penetration rate growing at an annual average of 15% [ 9 ]. In 1986, Davis introduced the Technology Acceptance Model, which posits that perceived usefulness and perceived ease of use are the primary determinants of user acceptance when encountering new technology [ 10 ]. Perceived usefulness refers to the extent to which users believe the technology enhances their work or learning efficiency, while perceived ease of use reflects the level of convenience users experience when interacting with the system. Together, these factors drive users' technology acceptance behavior. The application of Large-Scale Artificial Intelligence Models to library information retrieval systems enables intelligent filtering and recommendation functions, assisting users in quickly accessing high-quality information and boosting their confidence in obtaining reliable resources. Users with strong information assessment capability can more effectively assess the accuracy and value of system outputs, allowing them to explore the system's potential functionalities more deeply. This enhances their perception of the system's usefulness. The system's efficient feedback further encourages these users to continuously improve their information verification and critical analysis skills during sustained use, creating a positive feedback loop of 'system use—strengthened information assessment capability—increased perceived usefulness—higher usage frequency'. Based on this, Hypothesis 1a is proposed: The use of AI-powered library information retrieval systems can enhance users' information assessment capabilities. From the perspective of perceived ease of use, AI-powered library information retrieval systems feature user-friendly interface designs and diverse intelligent services, significantly lowering the operational barriers for users and enhancing their perception of the system's ease of use. Users with higher digital literacy can more quickly master the system's complex interaction logic, enabling them to utilize its functions more efficiently for retrieval and other tasks. This positive user experience further increases their frequency of system use [ 11 ]. The process of frequent system interaction inherently involves users accumulating digital skills through practice, including data processing, tool operation, and algorithm understanding and application, thereby contributing to an overall improvement in their digital literacy [ 12 ]. Accordingly, Hypothesis 1b is proposed: The use of AI-powered library information retrieval systems can enhance users' digital literacy. In 2021, Davy Ng and colleagues proposed an AI literacy framework encompassing four dimensions: 'Understanding AI', 'Applying AI', 'Evaluating and Creating AI', and 'AI Ethics' [ 13 ]. This framework is well-suited for examining how the application of Large-Scale Artificial Intelligence Models to library information retrieval systems can enhance ESD. Within the 'Understanding AI' dimension, university libraries deploying localized Large-Scale Artificial Intelligence Models provide users with a foundation for comprehending the underlying logic of AI technologies. By leveraging information retrieval systems powered by Large-Scale Artificial Intelligence Models, libraries educate users about the data processing mechanisms of these models, strengthening their foundational understanding of privacy and security awareness and establishing a cognitive basis for privacy protection. During the application phase of AI-powered library information retrieval systems, the system processes search commands and identifies personal preferences based on user inputs. Users with heightened privacy and security awareness are more likely to scrutinize the compliance of data collection, storage, and usage, adopting standardized operational behaviors to mitigate data breach risks. They also provide ongoing feedback regarding data security, model safety, and privacy protection needs, prompting libraries to enhance data governance and reinforcing users' awareness of digital ecosystem sustainability. In the 'Evaluating and Creating AI' phase, users' privacy and security awareness is further deepened. When assessing the quality, compliance, and safety of search results generated by Large-Scale Artificial Intelligence Models, privacy and security become critical evaluation metrics. Upon identifying potential data infringements or privacy risks in search outputs, users are prompted to critically examine the system's data ethics, continuously broadening and deepening their privacy and security awareness [ 14 ]. By leveraging the deployment and application of Large-Scale Artificial Intelligence Models, universities integrate privacy protection and data security into technical ethics education programs. Through targeted lectures, technical training, and classroom instruction, knowledge about data sovereignty and privacy is imparted to users, simultaneously enhancing their privacy and security awareness and advancing ESD. Based on this, Hypothesis 1c is proposed: The use of AI-powered library information retrieval systems can enhance users' privacy and security awareness. 2.2. Education for Sustainable Development The Sustainable Development Goals (SDGs) 4.7 represent a core component of the United Nations sustainable development agenda, with a particular focus on the quality and purpose of education. It's closely intertwined with ESD, drawing its vision from the United Nations Decade of ESD (2005–2014) and the subsequent Global Action Programme (2015–2019). Specifically, SDG 4.7 aims to ensure that by 2030, all learners acquire the knowledge and skills necessary for sustainable development. This is achieved through education on sustainable development, sustainable lifestyles, human rights, and gender equality, while promoting a culture of peace and non-violence, fostering global citizenship, and recognizing cultural diversity and its contributions to sustainable development [ 15 ]. ESD embodies an ethic of 'care' and 'concern', aimed at equipping learners with the knowledge, skills, and mindset to act for the planet by connecting their minds, hands, and hearts. It reorients education to enable learners of all ages to take action for environmental integrity, economic viability, cultural diversity, and a more equitable society. ESD emphasizes interdisciplinarity, fostering systems thinking to address the interconnected complexities of contemporary global challenges. From the perspective of the Technology Acceptance Model, users' acceptance of Large-Scale Artificial Intelligence Models in library information retrieval systems hinges on their perceptions of the technology's usefulness and ease of use. Users with strong information assessment capability can assess and filter search results, identifying valuable content, which encourages more active engagement with the system for learning and enhances their ESD. Users with high digital literacy are adept at operating information retrieval systems, fully leveraging the capabilities of Large-Scale Artificial Intelligence Models to efficiently access needed information. This positive experience of ease of use promotes frequent system use, increasing exposure to ESD content and ultimately strengthening its educational impact. The AI literacy framework emphasizes the knowledge, skills, and attitudes that individuals should possess in AI-driven environments. Users with privacy and security awareness understand concepts such as data security and information protection, enabling them to identify potential risks in acquiring and using information on ESD. This awareness helps prevent disruptions to educational activities caused by data breaches or misuse, ensuring the continuity and effectiveness of learning. Additionally, this awareness encourages users to acquire secure operational skills, such as encryption and authentication, ensuring that data and personal privacy are adequately protected when utilizing Large-Scale Artificial Intelligence Models in library information retrieval systems to access educational resources for sustainable development. This creates a secure environment conducive to in-depth learning [ 16 ]. From an attitudinal perspective, privacy and security awareness fosters a sense of responsibility among users to uphold information ethics and maintain a healthy digital ecosystem. Consequently, in engaging with ESD, users not only focus on their own knowledge acquisition but also actively contribute to building a fair, healthy, and orderly digital education environment, thereby supporting the realization of educational goals for sustainable development [ 16 ]. Based on the foregoing analysis, the following hypotheses are proposed: Hypothesis a: Users' information assessment capability is positively correlated with ESD. Hypothesis b: Users' digital literacy is positively correlated with ESD. Hypothesis c: Users' privacy and security awareness is positively correlated with ESD. 2.3. Application of Large-Scale Artificial Intelligence Models in Library Information Retrieval Systems and Education for Sustainable Development Current academic research on ESD primarily focuses on curriculum design, pedagogical approaches, technology integration, student engagement, and evaluation and policy support [ 17 – 21 ]. Literature on the application of technology in ESD aligns with theoretical frameworks and provides a robust foundation for this study. Applications of artificial intelligence in academia and education encompass educational support, constructive feedback, assessment, customized curricula, personalized career guidance, and mental health support [ 19 ]. Research indicates that Large-Scale Artificial Intelligence Models can significantly enhance valuable learning outcomes for higher education students [ 22 – 23 ]. This offers concrete practical references and theoretical grounding for advancing ESD through the integration of Large-Scale Artificial Intelligence Models. Although most studies support the notion that Large-Scale Artificial Intelligence Models enhance information access and learning efficiency, thereby promoting ESD, some controversies persist. Contrasting research highlights that ethical issues in AI technologies may undermine educational outcomes. Appleton argues that AI systems must address biases and privacy concerns to ensure their legitimate and compliant application in education, as failure to do so could negatively impact the learning environment of ESD [ 24 ]. Similarly, Fu emphasizes that excessive reliance on AI by users may diminish critical thinking skills, posing a challenge to the systemic objectives of ESD [ 25 ]. Based on the preceding analysis, the following hypotheses are proposed: Hypothesis a: The use of AI-powered library information retrieval systems is positively correlated with ESD. Sustainable development theory emphasizes the harmonious advancement of economic, social, and environmental dimensions, with education serving as a critical pathway to achieving these goals. Within library information retrieval systems, users' information assessment capability enables them to filter high-quality educational resources aligned with sustainable development principles, fostering greater awareness of sustainability issues and cultivating appropriate values and a sense of responsibility. Digital literacy empowers users to leverage Large-Scale Artificial Intelligence Models to bridge information gaps, accessing educational resources for sustainable development and promoting equitable sharing, thereby enhancing the inclusivity of ESD. Privacy and security awareness fosters a healthy digital ecosystem for improving ESD, ensuring the lawful and ethical use of information and facilitating the orderly conduct of educational activities. Together, these factors synergistically advance the objectives of ESD. Similarly, Lo underscores the need for libraries deploying Large-Scale Artificial Intelligence Models to simultaneously enhance the AI literacy of users and other stakeholders [ 26 ]. This involves engaging librarians in the implementation of AI, developing AI usage guidelines, and advocating that academic libraries proactively embrace AI to ensure its effective, ethical, and responsible applications in library services and operations. Based on this, the following Hypothesis is proposed: Hypothesis b: Users' information assessment capability, digital literacy, and privacy and security awareness partially moderate the relationship posited in Hypothesis 3a. 3. Materials and Methods 3.1. Sample selection and data sources This study collects data through a survey questionnaire. Informed consent was obtained from all participants and / or their legal guardians. This study complies with ethical standards and is approved by the Ethics Committee of School of Marxism, Harbin University of Science and Technology. Participants' written informed consents are obtained. All experiments were performed in accordance with relevant guidelines and regulations. In May 2025, a pre-test and pilot study was conducted with 48 students to assess the validity and reliability of the questionnaire. This preliminary survey helped evaluate the relevance, clarity, and comprehensiveness of the included scales, ensuring their suitability for the target population. In June 2025, a formal survey was administered to 315 respondents via the Wenjuanxing platform. The purpose of the questionnaire was to evaluate the pedagogical implications of applying Large-Scale Artificial Intelligence Models to library information retrieval systems, grounded in the theoretical framework of AI literacy. The survey was structured around the AI literacy framework and the United Nations Sustainable Development Goal 4 (SDG 4) framework, encompassing eight categories: understanding and comprehending AI; using and applying AI; evaluating and creating AI; AI ethics; digital literacy; cybersecurity awareness; educational equity and learning outcomes; and educational opportunities, skills, and sustainable development. No personal information was collected during the survey, ensuring complete anonymity. The collected data were securely stored and kept confidential to mitigate any risks to respondents' interests. Participants included a small number of high school students, as well as undergraduate and postgraduate students. Accordingly, the study employed a non-probability purposive sampling method to investigate the experiences of higher education students using Large-Scale Artificial Intelligence Models in library information retrieval systems. Etikan notes that non-probability purposive sampling is one of the most common techniques for selecting respondents with knowledge or experience relevant to a specific research question [ 27 ]. Therefore, selecting university students pursuing bachelor's, master's, and doctoral degrees was appropriate, as they represent a significant segment of higher education and have either experienced or are currently engaging with digital learning environments facilitated by Large-Scale Artificial Intelligence Models in library information retrieval systems. This non-probability sampling technique enabled researchers to deliberately select respondents with specific characteristics, ensuring the integrity and relevance of the collected data [ 28 ]. Data processing and analysis were conducted using SPSS, with subsets of the original dataset extracted based on specific criteria for further analysis. After excluding data from high school students and invalid responses, 205 valid responses remained for analysis. 3.2. Variable Definitions 3.2.1. Dependent variable The dependent variable is ESD. It is measured by averaging the scores of items in the sixth section (Educational Equity and Learning Outcomes) and the seventh section (Educational Opportunities, Skills, and Sustainable Development) of the questionnaire. Higher values indicate a greater level of ESD. 3.2.2 Independent variable The independent variable is the ability to use library information retrieval systems, assessed through AI usage capability. This is calculated by averaging the scores of all items in the second section (Using and Applying AI) of the questionnaire. Higher values reflect stronger proficiency in using library information retrieval systems. 3.2.3 Mediating variables This study selects digital literacy, information assessment capability, and privacy and security awareness as mediating variables to further analyze the pathways through which the application of Large-Scale Artificial Intelligence Models in library information retrieval systems impacts ESD. Digital literacy and information assessment capability are measured by averaging the scores of items in the fourth section (Digital Literacy) and the third section (AI Ethics and Assessment) of the questionnaire, respectively. Privacy and security awareness is assessed based on responses to a single questionnaire item, specifically the level of agreement with the statement, 'I believe it is very necessary to protect personal data in library AI systems' (rated on a scale from 1 = Strongly Disagree to 5 = Strongly Agree). The resulting value reflects the individual's emphasis on data privacy. 3.2.4 Control variables Drawing on existing literature, this study incorporates the following control variables to ensure the robustness of the findings: Gender: This refers to the biological sex of the participants and is used to explore potential differences in behavioral preferences and acceptance levels across genders. Variations in information retrieval habits and learning styles between genders may influence the accuracy of the study's results. Age: This represents the age group of the participants and is included to assess whether individuals in different age brackets exhibit varying responses when using library information retrieval systems embedded with Large-Scale Artificial Intelligence Models [ 29 ]. Age differences are often associated with disparities in learning capabilities and technology acceptance, which can significantly impact the study's conclusions. Grades: This indicates the educational stage of the students (e.g., undergraduate’s, master's, or doctoral level) and is used to analyze whether differences in knowledge reserves and learning objectives among students at various grades lead to differing outcomes when utilizing library information retrieval systems. Subject Background: This refers to the participants' academic fields [ 30 ]. It is included to investigate whether groups with different academic backgrounds produce varied results when accessing ESD resources through library information retrieval systems, due to differences in professional thinking and information needs. AI Education Level: This measures the extent to which participants have received education related to artificial intelligence [ 31 ]. It is used to evaluate whether individuals with varying levels of AI education exhibit different outcomes when using library information retrieval systems, influenced by their familiarity with AI technologies and application proficiency. AI education level significantly affects respondents' operational proficiency and system utilization efficiency, making it a critical factor. Table 1 presents the definitions of the main variables in the research analysis. Table 1 Definitions and measurement methods of key variables. Variable Type Variable Name notation Definition and Measurement Dependent Variable Education for Sustainable Development sustEdu Calculate the average scores of all items in Section 6 (Educational Equity and Learning Outcomes) and Section 7 (Educational Opportunities, Skills, and Sustainable Development) of the questionnaire. Independent Variable AI Usage capability aiSkill Calculate the average scores of all items in the dimension of Section 2 (Use and Application of AI) of the questionnaire. Mediating Variables Digital Literacy digLit Calculate the average scores of all items in the dimension of Section 4 (Digital Literacy) of the questionnaire. Information Assessment Capability infoAssess Calculate the average scores of all items in the dimension of Section 3 (AI Ethics and Assessment) of the questionnaire. Privacy and Security Awareness pvtSec Measure the degree of agreement with the statement 'I believe it is very necessary to protect personal data in library AI systems' (1 = strongly disagree, 5 = strongly agree) among respondents, using it as a proxy variable for individuals' emphasis on data privacy. Control Variables Gender Gender 'Male' = 0; 'Female' = 1. Grade Grade 'Freshman' = 1; 'Sophomore' = 2; 'Junior' = 3; 'Senior' = 4; 'Graduate (Master's/PhD)' = 5. AI Education Level aiEduLevel 'No exposure' = 1; 'Basic concept learning (less than 10 hours)' = 2; 'Systematic course learning (10–30 hours)' = 3; 'Professional application training (more than 30 hours)' = 4. Province prov Classify Chinese provinces into 'developed regions' and 'developing regions' based on their level of economic development, assigning a value of 1 to 'developed regions' and a value of 2 to 'developing regions'. Subject Background subjBkg 'Humanities and Social Sciences' = 1; 'Natural Sciences' = 2. 3.2.5. Model construction To analyze the impact of applying Large-Scale Artificial Intelligence Models to library information retrieval systems on ESD, Model (1) is constructed as follows: sustEdui = β 0 + β 1 ∙aiSkilli + β 2 ∙Genderi + β 3 ∙Gradei + β 4 ∙aiEduLeveli + β 5 ∙provi +β 6 ∙subjBkgi+ϵ i (1) To further investigate the mechanistic roles of digital literacy, information assessment capability, and privacy and security awareness, Model (2) is constructed as follows: sustEdui = γ 0 + γ 1 ∙aiSkilli + γ 2 ∙dgLiti + γ 3 ∙infoAssessi + γ 4 ∙pvtSeci+Controls + u i (2) In Models 1 and 2, sustEdu represents ESD and serves as the dependent variable. The independent variable aiSkill denotes the respondents' proficiency in using library AI information retrieval systems. The mediating variables digLit , infoAssess , and pvtSec represent digital literacy, information assessment capability, and privacy and security awareness, respectively. The control variables Gender , Grade , aiEduLevel , prov , and subjBkg correspond to gender, academic year, AI education level, province, and subject background, respectively, and are collectively denoted as Controls in Model 2. The subscript i indicates individual respondents. The terms ϵ i and u i represent the random error terms for Model 1 and Model 2, respectively. 3.3. Common Method Variance Analysis Data were collected through a survey questionnaire, with respondents primarily completing the questionnaire based on their subjective perceptions, which may introduce common method variance. To address this, we first employed Harman's single-factor test to assess the extent of common method variance in the data after collection. The results indicated that the maximum factor variance explained was 27.192%, which is below the critical threshold of 40%. This suggests that the issue of common method variance in the data was effectively controlled. 4. Results and Discussion 4.1. Demographics statistics Table 2 Demographics statistics. Variable Type Frequency Percent Gender Female 111 54.1 Male 94 45.9 Grade Freshman 36 17.6 Sophomore 54 26.3 Junior 57 27.8 Senior 46 22.4 Graduate (Master/PhD) 12 5.9 aiEduLevel Never touched 23 11.2 Basic concept learning (less than 10 class hours) 125 61 Systematic course learning (10–30 class hours) 53 25.9 Professional application training (more than 30 class hours) 4 2 prov developed regions 84 41 developing regions 121 59 subjBkg Humanities and Social Sciences 114 55.6 Natural Sciences 91 44.4 A total of 315 respondents participated in our questionnaire survey, with 205 valid samples remaining after data cleaning. Table 2 presents the demographic results. Of the respondents, 111 were female, and 94 were male. The majority were undergraduate students, accounting for 94.1% of the sample. Additionally, 61% of respondents had received formal education on basic concepts related to artificial intelligence. Regarding the respondents' provinces and subject backgrounds, the distributions were relatively balanced. Specifically, 41% of respondents were from economically developed regions, and 55.6% had a subject background in the humanities and social sciences. Table 3 Correlation Analysis. sustEdu aiSkill infoAssess digLit pvtSec prov Gender Grade subjBkg aiEduLevel sustEdu 1 aiSkill 0.688*** 1 infoAssess 0.561*** 0.560*** 1 digLit 0.763*** 0.826*** 0.621*** 1 pvtSec 0.403*** 0.292*** 0.256*** 0.334*** 1 prov 0.105 0.013 0.062 0.053 -0.001 1 Gender 0.139** 0.098 -0.025 0.141** 0.113 -0.010 1 Grade 0.171** 0.173** 0.118* 0.133* -0.051 -0.025 0.011 1 subjBkg 0.006 -0.005 0.035 0.033 0.018 0.145** -0.163** 0.024 1 aiEduLevel 0.238*** 0.246*** 0.296*** 0.306*** 0.030 0.024 0.022 0.094 0.109 1 The study employed Pearson's correlation test to examine the relationships among the variables, with the results presented in Table 3 . The correlation coefficient between aiSkill and sustEdu was 0.688, significant at the 1% level (denoted as ***). This indicates that the application of Large-Scale Artificial Intelligence Models in library information retrieval systems has a positive impact on ESD, providing preliminary support for Hypothesis 3a. The correlation coefficients between aiSkill and infoAssess , digLit , and pvtSec were 0.560, 0.826, and 0.292, respectively, all significant at the 1% level (denoted as ***). These findings suggest positive correlations between the application of Large-Scale. Artificial Intelligence Models in library information retrieval systems and information assessment capability, digital literacy, and privacy and security awareness, providing preliminary support for Hypotheses 1a, 1b, and 1c. Additionally, the correlation coefficients between infoAssess , digLit , pvtSec , and sustEdu were 0.561, 0.763, and 0.403, respectively, all significant at the 1% level (denoted as ***). This indicates strong positive correlations between information assessment capability, digital literacy, privacy and security awareness, and ESD, providing preliminary support for Hypotheses 2a, 2b, and 2c. 4.2. Reliability and validity testing of the questionnaire Table 4 Validity and reliability of constructs. Variable Item Measurements α KMO Bartlett’s Test of Sphericity approximate chi-square Degrees of Freedom Significance aiSkill 6 0.721 0.800 203.146 15 0.000 sustEdu 12 0.753 0.818 341.697 66 0.000 infoAssess 6 0.700 0.767 182.319 15 0.000 digLit 6 0.714 0.803 181.905 15 0.000 The study utilized scales for certain variables, and thus, reliability and validity tests were conducted on these scales, with results presented in Table 4 . Reliability Testing: The internal consistency of the scales was evaluated using Cronbach's α coefficient. The α values for all variables' scales exceeded 0.7, indicating that the measurements in this study possess high internal consistency [ 32 ]. Validity Testing: First, in terms of content validity, the initial measurement instrument was developed strictly based on the theoretical frameworks of artificial intelligence literacy and sustainable development. The questionnaire was refined and finalized through structured interviews with scholars in the same research field and pre-testing, ensuring strong content validity. Second, in terms of construct validity, the KMO test and Bartlett's test of sphericity were employed to assess whether the measurement items for each variable were suitable for factor analysis [ 33 – 34 ]. The KMO values for all scales exceeded the standard threshold of 0.7, and the P-values from Bartlett's test for the four variables were less than 0.001, indicating statistical significance. These results confirm the presence of common factors among the measurement item matrices, making them suitable for factor analysis and verifying construct validity. 4.3 Multicollinearity analysis Table 5 Statistical table of variance inflation factors. VIF 1/VIF digLit 3.795 0.263 aiSkill 3.251 0.308 infoAssess 1.734 0.577 pvtSec 1.163 0.86 aiEduLevel 1.151 0.869 Gender 1.082 0.924 subjBkg 1.069 0.935 Grade 1.049 0.953 prov 1.03 0.971 Mean VIF 1.703 . To prevent high correlations among the explanatory and control variables, which could lead to biased model estimates and to enhance the stability and explanatory power of the model, a multicollinearity test was conducted on the independent and control variables. The results are presented in Table 5 . Table 5 displays the Variance Inflation Factor (VIF) and its reciprocal (1/VIF) for different variables. VIF is an indicator used in multiple regression analysis to detect multicollinearity. Multicollinearity refers to high correlations among the independent variables in a model, which can compromise the stability of the regression model and the accuracy of the coefficients of the independent variables. Typically, a VIF value exceeding 10 or 5 indicates a serious multicollinearity issue. The VIF values for all variables reported in Table 5 are well below this threshold, suggesting that there is minimal multicollinearity among the independent and control variables. This indicates that the selection of variables is reasonable. 4.4. Baseline regression result Table 6 Model regression results. (1) (2) sustEdu sustEdu aiSkill 0.529*** 0.503*** (13.51) (12.34) prov 0.076* (1.90) Gender 0.057 (1.44) Grade 0.018 (1.03) subjBkg -0.001 (-0.03) aiEduLevel 0.041 (1.30) _cons 1.997*** 1.817*** (12.54) (10.15) N 205 205 R² 0.474 0.496 Note: (1) *, **, and *** mean statistical significance at the 10%, 5%, and 1% level, respectively.* p < 0.1, ** p < 0.05, *** p < 0.0; (2) t statistics in parentheses; In Table 6 , Column (1) presents the linear regression model without control variables, while Column (2) includes control variables. The regression results confirm the Hypothesis proposed in the study, namely that the application of Large-Scale Artificial Intelligence Models in library information retrieval systems has a positive impact on ESD. In the model without control variables, the coefficient is 0.529, and after introducing control variables, the coefficient is 0.503, both significant at the 1% level, thus validating Hypothesis 3a. Additionally, the province variable, as a control variable, also has a significant effect on ESD (coefficient = 0.076, p < 0.1), with this effect being more pronounced in developing regions. Furthermore, the improved model fit underscores the importance of control variables in reducing bias and enhancing the explanatory power of the model. 4.5. Mediation effect analysis Table 7 Model regression result ( infoAssess ). (1) (2) (3) sustEdu infoAssess sustEdu aiSkill 0.515*** 0.592*** 0.403*** (12.98) (9.47) (8.81) prov 0.076* 0.056 0.066* (1.90) (0.88) (1.71) Gender 0.058 -0.082 0.074* (1.46) (-1.30) (1.92) Grade 0.019 0.010 0.017 (1.10) (0.37) (1.03) subjBkg 0.005 0.018 0.001 (0.12) (0.28) (0.04) infoAssess 0.189*** (4.40) _cons 1.844*** 1.349*** 1.588*** (10.35) (4.81) (8.82) Sobel test 0.102*** (z = 3.790) Goodman-1 test 0.102*** (z = 3.811) Goodman-2 test 0.102*** (z = 3.770) Indirect effect coefficient 0.102*** (z = 3.790) Direct effect coefficient 0.401*** (z = 8.709) Total effect coefficient 0.503*** (z = 12.336) Proportion of intermediary effects 0.203 Bootstrap (ind_eff) 0.102*** [0.052,0.152] Bootstrap (dir_eff) 0.401*** [0.286,0.516] N 205 205 205 R² 0.491 0.324 0.537 Table 7 presents the regression results considering the mediating variable infoAssess . Prior to including the mediating variable infoAssess , the coefficient of aiSkill on sustEdu was 0.515 (p < 0.01). After introducing the mediating variable, the coefficient decreased to 0.403 (p < 0.01), still indicating a significant positive effect, but with reduced effect size. This suggests that infoAssess partially explains the relationship between aiSkill and sustEdu . The Sobel test and Goodman test confirmed the significance of the mediation effect of infoAssess . Additionally, Bootstrap analysis further validated this finding, with a 95% confidence interval [0.286, 0.516] that does not include 0, indicating that the mediating variable's role in transmitting the effect from the independent variable to the dependent variable is stable and not due to random error or spurious association. Thus, aiSkill indirectly promotes sustEdu by enhancing infoAssess . Overall, the mediation effect of infoAssess accounts for approximately 20% of the total effect (0.203), indicating that information assessment capability explains about 20% of the relationship between the application of Large-Scale Artificial Intelligence Models in library information retrieval systems and ESD. This suggests that merely implementing Large-Scale Artificial Intelligence Models in library information retrieval systems is insufficient to achieve education goals for sustainable development; it is also necessary to concurrently strengthen individuals' information assessment capabilities. Table 8 Model regression result ( digLit ). (1) (2) (3) sustEdu digLit sustEdu aiSkill 0.515*** 0.815*** 0.136** (12.98) (20.37) (2.21) prov 0.076* 0.036 0.060* (1.90) (0.89) (1.67) Gender 0.058 0.068* 0.027 (1.46) (1.68) (0.75) Grade 0.019 -0.004 0.021 (1.10) (-0.25) (1.37) subjBkg 0.005 0.043 -0.015 (0.12) (1.07) (-0.43) digLit 0.465*** (7.47) _cons 1.844*** 0.567*** 1.580*** (10.35) (3.16) (9.78) Sobel test 0.366*** (z = 6.842) Goodman-1 test 0.366*** (z = 6.850) Goodman-2 test 0.366*** (z = 6.834) Indirect effect coefficient 0.366*** (z = 6.842) Direct effect coefficient 0.136*** (z = 2.201) Total effect coefficient 0.503*** (z = 12.336) Proportion of intermediary effects 0.729 Bootstrap (ind_eff) 0.366*** [0.255,0.478] Bootstrap (dir_eff) 0.136* [-0.009,0.281] N 205 205 205 R² 0.491 0.690 0.603 Table 8 presents the regression results considering the mediating variable digLit . Similarly, after introducing digLit , the effect of aiSkill on sustEdu changed, with the coefficient decreasing from 0.515 to 0.136, and the significance level shifting from significant at the 1% level to significant at the 5% level. This preliminary finding suggests that digLit mediates the relationship between aiSkill and sustEdu . Subsequent Sobel and Goodman tests confirmed the significance of the mediation effect of digLit (z = 6.842–6.850, p < 0.01), with a mediation effect coefficient of 0.366. Bootstrap analysis further validated this result, with a 95% confidence interval [0.255, 0.478] that does not include 0, confirming that aiSkill indirectly influences sustEdu through digLit . Overall, the total effect coefficient is 0.503 (z = 12.336, p < 0.01), with the mediation effect of digLit accounting for approximately 73% (0.729) of the total effect, which is substantially higher than the mediation effect of infoAssess . Table 9 Model regression result ( pvtSec ). (1) (2) (3) sustEdu pvtSec sustEdu aiSkill 0.515*** 0.347*** 0.463*** (12.98) (4.41) (11.64) prov 0.076* -0.015 0.079** (1.90) (-0.18) (2.05) Gender 0.058 0.105 0.042 (1.46) (1.33) (1.11) Grade 0.019 -0.052 0.027 (1.10) (-1.55) (1.62) subjBkg 0.005 0.045 -0.002 (0.12) (0.56) (-0.05) pvtSec 0.150*** (4.40) _cons 1.844*** 3.215*** 1.360*** (10.35) (9.10) (6.71) Sobel test 0.055*** (z = 3.148) Goodman-1 test 0.055*** (z = 3.188) Goodman-2 test 0.055*** (z = 3.109) Indirect effect coefficient 0.055*** (z = 3.148) Direct effect coefficient 0.448*** (z = 10.979) Total effect coefficient 0.503*** (z = 12.336) Proportion of intermediary effects 0.109 Bootstrap (ind_eff) 0.055*** [0.020,0.089] Bootstrap (dir_eff) 0.448*** [0.346,0.550] N 205 205 205 R² 0.491 0.104 0.537 Table 9 presents the regression results considering the mediating variable pvtSec . Similarly, after introducing pvtSec , the coefficient of aiSkill on sustEdu changed from 0.515 to 0.463, with no change in statistical significance. The Sobel and Goodman tests confirmed the significance of the mediation effect of pvtSec (z = 3.148–3.109, p < 0.01), with a mediation effect coefficient of 0.055. Bootstrap analysis further validated this finding, with a 95% confidence interval [0.020, 0.089] that does not include 0, confirming that aiSkill indirectly influences sustEdu through pvtSec . Similarly, calculations indicate that the mediation effect of pvtSec accounts for approximately 11% (0.109) of the total effect, suggesting a relatively weak but statistically significant mediating role. The examination of the mediation effects further validates the hypotheses proposed in the study (1a–3b). All mediation effects were significant, indicating that the three variables ( infoAssess , digLit , and pvtSec ) each play a mediating role in the pathway through which aiSkill influences sustEdu . Among these, digLit exhibited the strongest mediation effect, accounting for approximately 73% of the total effect, while pvtSec had the weakest mediation effect, contributing about 11% to the total effect. Further analysis reveals that infoAssess supports the knowledge acquisition and application for sustainable education by enhancing individuals' critical evaluation capabilities of information. DigLit facilitates learners' effective participation in sustainable education practices within digital environments by improving proficiency in operating digital tools. Although the mediation effect of pvtSec is relatively smaller, it still contributes to the ethical foundation of sustainable education by increasing awareness of data privacy and security. These findings not only confirm the driving role of technology use in capability development as posited by the Technology Acceptance Model but also align with the integrated perspective of the artificial intelligence literacy framework and sustainable development theory. They demonstrate that the mediating variables establish critical transmission pathways between technology and educational outcomes. 4.6. Robustness test Table 10 Robustness test. (1) (2) sustEdu sustEdu aiSkill 0.503*** (12.34) aiFreq 0.088*** (2.98) prov 0.076* 0.081 (1.90) (1.55) Gender 0.057 0.106** (1.44) (2.05) Grade 0.018 0.035 (1.03) (1.53) subjBkg -0.001 0.006 (-0.03) (0.11) aiEduLevel 0.041 0.105** (1.30) (2.56) _cons 1.817*** 3.352*** (10.15) (21.13) N 205 205 R² 0.496 0.146 Note: t statistics in parentheses, * p < 0.1, ** p < 0.05, *** p < 0.01 The study conducted a robustness test by replacing the explanatory variable, substituting the proficiency in using library information retrieval systems ( aiSkill ) with the frequency of using library AI information retrieval systems ( aiFreq ). The results are presented in Table 10 . Specifically, the original variable aiSkill (representing the use of Large-Scale Artificial Intelligence Models in library information retrieval systems) was replaced with aiFreq (representing the frequency of using library AI information retrieval systems), and regression analysis was performed to assess the robustness of the model. The results indicate that the direct effect coefficient of aiSkill on sustEdu in Model (1) was 0.503 (p < 0.01). In Model (2), after replacing the variable, the effect of aiFreq on sustEdu remained significant, with a coefficient of 0.088 (p < 0.01), demonstrating that the frequency of use has a significant positive impact on ESD. The robustness test results suggest that, although the substitution of variables led to changes in the magnitude of the effect, the positive impact of aiSkill on sustEdu remained consistent across different measurement dimensions, confirming the reliability of the model results. This finding supports the stability of the research hypotheses and indicates that usage frequency and usage proficiency may contribute differently to explaining the effects on sustainable education, warranting further exploration in future research. 4.7. Heterogeneity test Table 11 Heterogeneity test. (1) developed regions (2) developing regions (3) Female (4) Male (5) Low AI Education Level (6) High AI Education Level sustEdu sustEdu sustEdu sustEdu sustEdu sustEdu aiSkill 0.605*** 0.449*** 0.386*** 0.594*** 0.471*** 0.556*** (8.51) (8.82) (6.60) (10.42) (9.60) (8.59) Gender 0.067 0.054 0.067 0.065 (1.13) (0.99) (1.33) (1.20) Grade -0.002 0.033 0.018 0.020 0.032 -0.021 (-0.07) (1.42) (0.77) (0.82) (1.50) (-0.83) subjBkg 0.003 0.012 0.026 -0.036 0.013 -0.041 (0.05) (0.23) (0.49) (-0.59) (0.26) (-0.71) aiEduLevel 0.067 0.009 0.082 0.012 (1.30) (0.21) (1.63) (0.30) prov 0.067 0.103 0.101** 0.009 (1.29) (1.64) (2.04) (0.16) _cons 1.467*** 2.198*** 2.236*** 1.519*** 1.897*** 2.033*** (5.20) (10.03) (9.01) (5.76) (8.65) (6.74) N 84 121 111 94 148 57 R² 0.577 0.453 0.377 0.587 0.449 0.606 The study further conducted a heterogeneity analysis to examine the varying effects of aiSkill on sustEdu across regions with different levels of economic development, groups of different genders, and groups with varying levels of artificial intelligence education. The results are presented in Table 11 . Heterogeneity Analysis Across Regions with Different Levels of Economic Development. To further explore the heterogeneity in the effect of aiSkill on sustEdu , the study first examined the differential effects across regions with varying levels of economic development. In regions with higher economic development (Column 1), the regression coefficient of aiSkill on sustEdu was 0.605 (p < 0.01), with an R² of 0.577. In contrast, in regions with lower economic development (Column 2), the coefficient decreased to 0.449 (p < 0.01), with an R² of 0.453. These results indicate that the promoting effect of aiSkill on sustainable development education is more pronounced in economically developed regions. In economically developed regions, well-established hardware infrastructure ensures the stable operation of educational applications related to aiSkill . Abundant data resources support the optimization of algorithms and the customization of personalized teaching plans. An open and innovative educational ecosystem, coupled with teachers' higher digital literacy, facilitates the rapid integration of aiSkill into sustainable development education classrooms. Additionally, strong societal demand for talent and ample policy and financial support further promote the deep integration of aiSkill and sustainable education. The convergence of these favorable conditions amplifies the positive impact of aiSkill on ESD. This heterogeneity reflects the adaptive differences in the application of Large-Scale Artificial Intelligence Models in library information retrieval systems across diverse socioeconomic contexts. It further suggests that future educational interventions should be tailored to regional characteristics to address these disparities effectively. Heterogeneity Analysis Across Gender Groups. The study also analyzed the heterogeneity in the effect of aiSkill on sustEdu across different gender groups. The results indicate that in the female group (Column 3), the regression coefficient of aiSkill on sustEdu was 0.386 (p < 0.01), with an R² of 0.377. In the male group (Column 4), the coefficient was 0.594 (p < 0.01), with an R² of 0.587. Although both effects are significant, the effect is slightly weaker in the female group, which may be related to differences in learning strategies and personality tendencies. Gender differences are influenced by sociocultural norms or educational support systems. To narrow this impact gap, the participation of females in artificial intelligence skills training should be further enhanced through targeted interventions. Heterogeneity Analysis Across Groups with Different Levels of AI Education. Additionally, the study examined the heterogeneity in the effect of aiSkill on sustEdu across groups with varying levels of artificial intelligence education. The results indicate that in the group with a low level of AI education (Column 5), the coefficient of aiSkill on sustEdu was 0.471 (p < 0.01), with an R² of 0.449. In contrast, in the group with a high level of AI education (Column 6), the coefficient increased to 0.556 (p < 0.01), with an R² of 0.606. The effect size in the high-level AI education group was significantly stronger than in the low-level group, reflecting the advantages of advanced education in deepening skills and broadening application scope, which enhances its promoting effect on sustainable development education. The increase in R² values indicates that the model has greater explanatory power in the high-level AI education group. This heterogeneity stems from the cumulative effect of education level, where high-level AI education not only enhances technical proficiency but also strengthens learners' ability to translate skills into educational outcomes in complex contexts. However, the effect in the low-level group, though weaker, remains significant, suggesting that basic AI skills still hold practical value at the initial stage. These findings recommend a tiered approach to AI education: for low-level groups, foundational application training should be strengthened, while high-level groups should receive support for advanced projects to maximize the overall impact of aiSkill on sustEdu . 5. Conclusions This study investigates the impact of applying Large-Scale Artificial Intelligence Models to library information retrieval systems on ESD, the pathways of this impact, and its heterogeneity across different gender groups, regions, and groups with varying levels of AI education. The Technology Acceptance Model, the artificial intelligence literacy framework, and sustainable development theory collectively form an integrated model to explain the effects on ESD. Compared to individual theories, the results demonstrate that the integrated model offers greater explanatory power. From the perspective of the Technology Acceptance Model, users' perceived usefulness and ease of use of library AI retrieval systems drive their usage behavior, which in turn influences their engagement with sustainable development education. Within the artificial intelligence literacy framework, information assessment capability, digital literacy, and privacy and security awareness serve as key mediators: information assessment capability enables users to accurately filter information relevant to ESD, digital literacy ensures efficient use of the system to access resources, and privacy and security awareness establishes a foundation of trust in system usage. Sustainable development theory, from a macro perspective, explains how education promotes the coordinated development of society, the environment, and the economy, providing direction for the impact pathways. Compared to a single theory, this integrated approach mitigates the Technology Acceptance Model's limited focus on users' intrinsic competencies, compensates for the artificial intelligence literacy framework's lack of macro-level value articulation, and addresses the shortcomings of sustainable development theory in elucidating micro-level pathways for technology application. 5.1.Recommendations Based on the findings that the application of Large-Scale Artificial Intelligence Models in library information retrieval systems can facilitate the achievement of educational goals for sustainable development, the study proposes the following recommendations: Policy Recommendations: Government and educational authorities should promote the localized deployment of Large-Scale Artificial Intelligence Models in higher education institutions and advance the digitization of university and local libraries through policy support and funding to enhance the realization of ESD objectives. The study demonstrates that the application of library AI retrieval systems significantly influences ESD through information assessment capability, digital literacy, and privacy and security awareness, with digLit accounting for 73% of the total mediation effect (Table 8 ). To this end, policies could draw on the European Union's Digital Education Action Plan (2021–2027) by allocating dedicated funds and implementing training programs to support the integration of AI technologies in education [ 35 ]. Educational policies and frameworks should emphasize the development of students' digital literacy and information assessment capabilities, which are critical for the interdisciplinary learning required in ESD [ 36 ]. Furthermore, the heterogeneity analysis reveals that the effect of aiSkill is stronger in economically developed regions, among males, and in groups with higher levels of AI education, suggesting that policies should optimize resource allocation by accounting for regional and individual differences. Additionally, stringent data privacy and security standards should be established to ensure compliance and legality in the application of Large-Scale Artificial Intelligence Models in educational settings, thereby enhancing users' trust in AI systems and improving usage willingness and educational outcomes. Educational Practice. Based on the research findings, higher education institutions should integrate knowledge about AI information retrieval systems, applications of Large-Scale Artificial Intelligence Models, and related literacy into curriculum design to foster students' sustainable development awareness and technical skills. The study demonstrates that the use of library AI information retrieval systems promotes sustEdu by enhancing digLit , infoAssess , and pvtSec . Educators should receive professional training to effectively guide students in using AI tools within library settings. Additionally, the evaluation systems for higher education should be updated to include students' abilities to address sustainable development issues using AI tools, thereby comprehensively assessing their learning outcomes. Furthermore, the robustness test indicates that the effect remains significant even after variable substitution, underscoring that educational practices should prioritize skill development rather than merely relying on usage frequency. This approach not only supports the enhancement of perceived usefulness and ease of use as outlined in the Technology Acceptance Model, but also aligns with the artificial intelligence literacy framework's goals of 'applying AI' and 'evaluating and creating AI'. International Experience. A recent report from the U.S. Department of Education, 'Artificial Intelligence and the Future of Teaching and Learning', provides valuable insights into the role of artificial intelligence in education and its implications for various stakeholders, including academic libraries. International experiences offer valuable references for the application of Large-Scale Artificial Intelligence Models in Chinese university libraries. In the United Kingdom, university libraries have enhanced research efficiency through AI tools, such as the AI recommendation system implemented by the University of London Library. This successful experience can serve as a model for Chinese higher education institutions. The European Union promotes AI literacy education through cross-national collaborative projects, emphasizing the cultivation of ethical awareness, which aligns with the findings in this study that Chinese universities need to strengthen pvtSec . In contrast, developing countries like India have expanded access to educational resources through low-cost AI solutions, such as voice assistants, offering a reference for less developed regions in China. By integrating these experiences, China should formulate localized strategies that balance technological advancement with educational equity to achieve sustainable development goals. 5.2. Limitations and future work Although this study systematically analyzed the promoting effect of library AI information retrieval system usage on ESD and its mediating mechanisms through the Technology Acceptance Model, the artificial intelligence literacy framework, and sustainable development theory, several limitations remain. First, the sample was primarily composed of students from Chinese higher education institutions, and the homogeneity in geographic and cultural backgrounds limits the generalizability of the findings. Future research could expand to include international multicenter samples to compare the impact of Large-Scale Artificial Intelligence Models usage in library information retrieval systems on ESD across diverse cultural contexts. Second, the study relied on self-reported data and cross-sectional analysis, which could not capture the dynamic changes in mediating variables resulting from long-term usage behavior. Future studies could employ longitudinal research or experimental designs to validate causal relationships. Third, the study did not thoroughly explore the technical characteristics or ethical issues of AI systems. Future research could incorporate technical evaluation frameworks to analyze their potential negative effects. Declarations Conflicts of Interest: The authors declare no competing interests. Funding: This research received no external funding. Author Contribution Conceptualization, Y.Z., Z.L., and C.W.; data curation, Y.Z. and Z.L.; formal analysis, Y.Z., Z.L., and C.W.; investigation, Y.Z., C.W., and Z.L.; methodology, Y.Z., C.W., and Z.L.; project administration, C.W.; resources, Y.Z., Z.L., and C.W.; supervision, C.W.; validation, Y.Z., Z.L., and C.W.; visualization, Z.L.; writing—original draft, Y.Z. and Z.L.; writing—review and editing, Y.Z., Z.L., and C.W.. All the authors have read and agreed to the published version of the manuscript. Data Availability The data presented in this study are available on request from the corresponding author due to questionnaire respondents' confidentiality. References United Nations General Assembly. Education for sustainable development in the framework of the 2030 Agenda for Sustainable Development: Report of the Director-General of UNESCO (A/78/219). (2023). Available at: https://docs.un.org/en/A/78/219 Giannini, S. Education for sustainable development—Roadmap. United Nations Educational, Scientific and Cultural Organization (2020). https://doi.org/10.54675/YFRE1448 Singun, A. Jr. Unveiling the barriers to digital transformation in higher education institutions: a systematic literature review. Discov Educ. 4 , 37 (2025). Avelar, A. B. A., Mitchell, S. F. & Sandes-Guimarães, L. V. Integrating sustainable development goals in management education: Impact on student knowledge, attitudes, and behaviors. International J. Manage. Education . 23 , 101116 (2025). Savec, V. F. & Jedrinović, S. The Role of AI Implementation in Higher Education in Achieving the Sustainable Development Goals: A Case Study from Slovenia. Sustainability 17, (2024). Goralski, M. A. & Tan, T. K. Artificial intelligence and sustainable development. Int. J. Manage. Educ. 18 , 100330 (2020). Garnier, M., Cussac, M., Delaune, C. & Espy, C. & Papon-Vidal, L. The Pivotal Role of Libraries in Sustainable AI Development. cjils 47, 137–149 (2024). Khan, R., Gupta, N., Sinhababu, A. & Chakravarty, R. Impact of Conversational and Generative AI Systems on Libraries: A Use Case Large Language Model (LLM). Science & Technol. Libraries . 43 , 319–333 (2024). International Federation of Cartographic Associations. The strategy of the International Federation of Cartographic Associations 2024–2029. (2024). Available at: https://www.lsc.org.cn/cns/html_pass/contents/1127/1848974105203838976.html Khan, R., Gupta, N., Sinhababu, A. & Chakravarty, R. Impact of Conversational and Generative AI Systems on Libraries: A Use Case Large Language Model (LLM). Science & Technol. Libraries . 43 , 319–333 (2024). Adewojo, A. A., Amzat, O. B. & Abiola, H. S. AI-powered libraries: enhancing user experience and efficiency in Nigerian knowledge repositories. Libr. Hi Tech. News . 42 , 12–16 (2025). Chen, S., Xue, Y. & Cui, X. Information literacy of college students from library education in smart classrooms: Based on big data exploring data mining patterns using Apriori algorithm. Soft. Comput. 28 , 3571–3589 (2024). Ng, D. T. K., Leung, J. K. L., Chu, S. K. W. & Qiao, M. S. Conceptualizing AI literacy: An exploratory review. Computers Education: Artif. Intell. 2 , 100041 (2021). Korobenko, D., Nikiforova, A. & Sharma, R. Towards a Privacy and Security-Aware Framework for Ethical AI: Guiding the Development and Assessment of AI Systems. in Proceedings of the 25th Annual International Conference on Digital Government Research 740–753Association for Computing Machinery, New York, NY, USA, (2024). 10.1145/3657054.3657141 United Nations Educational, Scientific and Cultural Organization. Target 4.7: Sustainable development and global citizenship. (2019). Available at: https://gem-report-2019.unesco.org/chapter/monitoring-progress-in-sdg-4/sustainable-development-and-global-citizenship-target-4-7/ Vettriselvan, R. Empowering Digital Education: The Future of Value-Based Learning in the Digital Era. in Empowering Value Co-Creation in the Digital Era 199–228IGI Global Scientific Publishing, (2025). 10.4018/979-8-3373-1742-7.ch009 Almaleh, A., Aslam, M. A., Saeedi, K. & Aljohani, N. R. Align My Curriculum: A Framework to Bridge the Gap between Acquired University Curriculum and Required Market Skills. Sustainability 11, (2019). Rajalakshmi, S. A. & Dynamic, L. J. N. R., P, V. Paul. NLP Enabled Chatbot for Rural Health Care in India. in Second International Conference on Computer Science, Engineering and Applications (ICCSEA) 1–6 (2022). 1–6 (2022). (2022). 10.1109/ICCSEA54677.2022.9936389 Alqahtani, T. et al. The emergent role of artificial intelligence, natural learning processing, and large language models in higher education and research. Res. Social Administrative Pharm. 19 , 1236–1242 (2023). Agostini, D. & Picasso, F. Large language models for sustainable assessment and feedback in higher education: Towards a Pedagogical and Technological Framework. Intelligenza Artificiale . 18 , 121–138 (2024). Mbakwe, A. B., Lourentzou, I., Celi, L. A., Mechanic, O. J. & Dagan, A. ChatGPT passing USMLE shines a spotlight on the flaws of medical education. PLOS Digit. Health . 2 , e0000205 (2023). Huang, H. et al. ChatGPT for shaping the future of dentistry: the potential of multi-modal large language model. Int. J. Oral Sci. 15 , 29 (2023). Al-Dokhny, A. et al. Can Multimodal Large Language Models Enhance Performance Benefits Among Higher Education Students? An Investigation Based on the Task–Technology Fit Theory and the Artificial Intelligence Device Use Acceptance Model. Sustainability 16 , 10780 (2024). Appleton, L. AI and academic libraries: what’s all the fuss about? New. Rev. Acad. Librariansh. 30 , 1–3 (2024). Fu, Y., Weng, Z., Wang, J. & Examining AI Use in Educational Contexts: A Scoping Meta-Review and Bibliometric Analysis. Int. J. Artif. Intell. Educ. 35 , 1388–1444 (2024). Lo, L. S. An initial interpretation of the U.S. Department of Education's AI report: Implications and recommendations for academic libraries. J. Acad. Librariansh. 49 (5), 102761 (2023). Etikan, I. Sampling and Sampling Methods. BBIJ 5, (2017). Bouncken, R. B., Czakon, W. & Schmitt, F. Purposeful sampling and saturation in qualitative research methodologies: recommendations and review. Rev Manag Sci (2025). Jia, J., Yuan, Z., Pan, J., McNamara, P. E. & Chen, D. Decision-making behavior evaluation framework for LLMs under uncertain context. Proceedings of the 38th International Conference on Neural Information Processing Systems (NIPS '24) 37, Article 3601, pp. 113360–113382 (2025). Grande, V., Kiesler, N. & Francisco, R. M. A. Student Perspectives on Using a Large Language Model (LLM) for an Assignment on Professional Ethics. in Proceedings of the 2024 on Innovation and Technology in Computer Science Education V. 1 478–484ACM, Milan Italy (2024). 10.1145/3649217.3653624 Kozov, V., Ivanova, B., Shoylekova, K. & Andreeva, M. Analyzing the Impact of a Structured LLM Workshop in Different Education Levels. Appl. Sci. 14 , 6280 (2024). Yadav, R. & Pathak, G. S. Determinants of Consumers’ Green Purchase Behavior in a Developing Nation: Applying and Extending the Theory of Planned Behavior. Ecol. Econ. 134 , 114–122 (2017). Chen, M. F. & Tung, P. J. Developing an extended Theory of Planned Behavior model to predict consumers’ intention to visit green hotels. Int. J. Hospitality Manage. 36 , 221–230 (2014). Chen, Y. C., Ho, C. C. & Kuo, S. M. Service Quality of and User Satisfaction with Non-State-Owned Academic Libraries in China: Integrating the Fuzzy Delphi Method with the Kano Approach. Sustainability 14 , 8506 (2022). European Commission. Digital Education Action Plan. (2021–2027). Available at: https://education.ec.europa.eu/focus-topics/digital-education/action-plan U.S. Department of Education, Office of Educational Technology. Artificial Intelligence and the Future of Teaching and Learning: Insights and Recommendations. (2023). Available at: https://event-cdn.baai.ac.cn/file/file-browser/hAZcnhiJsJaGdyBnFhZjw3Wa7fPpGZ5D.pdf Additional Declarations No competing interests reported. Cite Share Download PDF Status: Under Review Version 1 posted Editorial decision: Revision requested 23 Mar, 2026 Reviews received at journal 06 Mar, 2026 Reviews received at journal 28 Feb, 2026 Reviewers agreed at journal 17 Feb, 2026 Reviewers agreed at journal 14 Feb, 2026 Reviews received at journal 14 Feb, 2026 Reviewers agreed at journal 14 Feb, 2026 Reviewers agreed at journal 13 Feb, 2026 Reviews received at journal 13 Feb, 2026 Reviewers agreed at journal 12 Feb, 2026 Reviewers invited by journal 12 Feb, 2026 Editor assigned by journal 10 Feb, 2026 Editor invited by journal 06 Feb, 2026 Submission checks completed at journal 05 Feb, 2026 First submitted to journal 05 Feb, 2026 You are reading this latest preprint version Research Square lets you share your work early, gain feedback from the community, and start making changes to your manuscript prior to peer review in a journal. As a division of Research Square Company, we’re committed to making research communication faster, fairer, and more useful. We do this by developing innovative software and high quality services for the global research community. Our growing team is made up of researchers and industry professionals working together to solve the most critical problems facing scientific publishing. Also discoverable on Platform About Our Team In Review Editorial Policies Advisory Board Help Center Resources Author Services Accessibility API Access RSS feed Manage Cookie Preferences © Research Square 2026 | ISSN 2693-5015 (online) Privacy Policy Terms of Service Do Not Sell My Personal Information {"props":{"pageProps":{"initialData":{"identity":"rs-8751190","acceptedTermsAndConditions":true,"allowDirectSubmit":false,"archivedVersions":[],"articleType":"Article","associatedPublications":[],"authors":[{"id":591289001,"identity":"96d1d7a0-332a-4c3c-8620-f980ddf3f25b","order_by":0,"name":"Yunping Zhao","email":"","orcid":"","institution":"Harbin University of Science and Technology","correspondingAuthor":false,"prefix":"","firstName":"Yunping","middleName":"","lastName":"Zhao","suffix":""},{"id":591289002,"identity":"d546add4-5342-48cd-ae4a-c26df7437227","order_by":1,"name":"Zhou Lan","email":"","orcid":"","institution":"University of Santo Tomas","correspondingAuthor":false,"prefix":"","firstName":"Zhou","middleName":"","lastName":"Lan","suffix":""},{"id":591289003,"identity":"bf1c83ea-78aa-473d-81d8-f215d55c65b6","order_by":2,"name":"Chengfei Wei","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAABAklEQVRIie3QsWrDMBCA4TMGZ7lGHc8E/AwHgtAhNK+iEPDkIWOnEjDES7xn6Et0adpNReDJJWvGhoKnDJ5KA4FWXkOrplsH/YMGoU+cBODz/cNEGGrdMuG4yDXAqNtTbhIX0WS3ml0ljJU9mp5BeINSYnsjmTI+kxgYEjJN1li/v6EagehZe3h0DGYgJbLkqSjXElUK8XLPQVn/TPoGKmJL5vXLw+DiwwBvMw6DhWMyEyxIdWSbNQNUnzD+jVyaMGTNJO3lkSUaun9wkjiPgt2cKYmX1TC+U1Okupk9lw4ixKY1x+Mtil7e0F5dJ6KY3r8eXG85DbtF/wH4fD6f75u+ABfGUF2ojwp2AAAAAElFTkSuQmCC","orcid":"","institution":"Nanjing University of Finance and Economics","correspondingAuthor":true,"prefix":"","firstName":"Chengfei","middleName":"","lastName":"Wei","suffix":""}],"badges":[],"createdAt":"2026-01-31 16:08:10","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-8751190/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-8751190/v1","draftVersion":[],"editorialEvents":[],"editorialNote":"","failedWorkflow":false,"files":[{"id":102822487,"identity":"f51e5307-7c3f-42f7-ae77-1eb63187efeb","added_by":"auto","created_at":"2026-02-17 07:56:42","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":1535008,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-8751190/v1/df60b3f7-b9bf-47df-acff-65898b7cb801.pdf"}],"financialInterests":"No competing interests reported.","formattedTitle":"Promoting Education for Sustainable Development through AI-Enhanced Library Information Retrieval Systems","fulltext":[{"header":"1. Introduction","content":"\u003cp\u003eEducation is a cornerstone of sustainable development, playing an indispensable role in advancing the global transition toward a sustainable future. In recent years, education for sustainable development (ESD) has been increasingly recognized as a vital means of achieving sustainability, with its contributions to quality education and lifelong learning gaining acknowledgment from national societies, major countries and regions, and other stakeholders.\u003c/p\u003e \u003cp\u003eESD was designated as one of the United Nations' Sustainable Development Goals in 2015. In January 2016, the 2030 Agenda for Sustainable Development explicitly called for inclusive and equitable quality education across various stages, including early childhood, primary, secondary, and higher education, as well as technical and vocational training. It urged countries to provide financial and technical support, develop policies and programs for high-quality distance learning in higher education, and encourage early student engagement with science, technology, engineering, and mathematics fields [\u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e1\u003c/span\u003e]. Education for Sustainable Development: Towards achieving the SDGs (ESD for 2030) is the theme of the third phase of UNESCO's global action plan for ESD. This plan was approved in September 2019 and officially launched in June 2020. Its accompanying document, Education for Sustainable Development: A Roadmap, highlights key areas, including but not limited to: the intersection of ESD and the sustainable development goals, transformative actions and ESD, structural issues and ESD, and technological advancements and ESD. Within the dimension of technological advancements and ESD, critical areas of exploration include the impact of artificial intelligence, the Fourth Industrial Revolution, and the pathways through which new technologies can empower sustainable education [\u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2\u003c/span\u003e]. These themes provide the broader context for this study.\u003c/p\u003e \u003cp\u003eThe widespread adoption of Large-Scale Artificial Intelligence Models and other digital technologies is driving the digital transformation of education [\u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e3\u003c/span\u003e]. Within this context, libraries, as critical hubs for knowledge dissemination and lifelong learning, play a pivotal role. The iterative enhancement of their information retrieval systems has a profound impact on advancing ESD. This form of education seeks to cultivate versatile individuals equipped with environmental awareness, social responsibility, and innovative capabilities [\u003cspan citationid=\"CR4\" class=\"CitationRef\"\u003e4\u003c/span\u003e]. As repositories of knowledge, the optimization of libraries' digital services directly influences the accessibility and dissemination efficiency of educational resources for sustainable development. To better meet the increasingly diverse academic needs of students and faculty, there is an urgent need to advance the intelligent upgrading of library information retrieval systems. Currently, AI technologies are experiencing rapid advancement. LLMs, as a groundbreaking achievement in the field of AI, are deeply permeating various industries. Trained on vast datasets, LLMs demonstrate exceptional capabilities in natural language understanding, generation, and knowledge association, enabling precise analysis of complex semantics and the identification of latent connections of different pieces of information. In information retrieval applications, Large-Scale Artificial Intelligence Models overcome the limitations of traditional retrieval systems that rely on keyword matching, offering greater accuracy in interpreting user needs and organizing knowledge resources. An increasing number of libraries are adopting Large-Scale Artificial Intelligence Models. The new generation of library information retrieval systems, leveraging the algorithmic advantages of these models, can rapidly locate vast online academic resources and digitized literature. Additionally, these systems can construct interdisciplinary knowledge graphs, accurately integrating research outcomes and practical case studies from the field of sustainable developments, including spanning environmental science, social governance, economic transformation, and other dimensions, thereby providing learners with a systematic knowledge framework. Furthermore, library information retrieval systems transcend geographical and temporal constraints, enabling educators and students in remote areas to access high-quality educational resources equitably, thus promoting the inclusive sharing of educational resources for sustainable development. Additionally, these systems can incorporate interactive learning modules, facilitating activities such as knowledge quizzes and case study discussions related to sustainable development themes. This enhances user engagement and fosters a widespread commitment to sustainability. Consequently, libraries are transitioning into intelligent knowledge service hubs, providing sustained momentum for achieving the goals of ESD.\u003c/p\u003e \u003cp\u003eCurrently, research exploring the connection between Large-Scale Artificial Intelligence Models and ESD is in its preliminary stages, yet significant gaps remain [\u003cspan citationid=\"CR5\" class=\"CitationRef\"\u003e5\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR6\" class=\"CitationRef\"\u003e6\u003c/span\u003e]. In particular, the impact of applying Large-Scale Artificial Intelligence Models in specific contexts on ESD remains underexplored. Existing research primarily focuses on the superficial impact of technological applications on knowledge acquisition models, with limited exploration of the long-term mechanisms through which the transformation of library information retrieval ecosystems, driven by Large-Scale Artificial Intelligence Models, shapes ESD [\u003cspan citationid=\"CR7\" class=\"CitationRef\"\u003e7\u003c/span\u003e]. Furthermore, when integrating AI literacy frameworks, there is a lack of detailed analysis regarding the differential impact pathways for users of varying genders, regions, and educational attainment levels. It remains unclear how the application of Large-Scale Artificial Intelligence Models in library information retrieval systems can precisely address the diverse needs of ESD. Addressing this research gap is critical for the decision-making and implementation of digital library initiatives.\u003c/p\u003e \u003cp\u003eThis study examines 'the impact of applying Large-Scale Artificial Intelligence Models to library information retrieval systems on ESD'. Through empirical methods, it analyzes the effects, pathways, and heterogeneity of these impacts across groups differentiated by gender, region, and levels of AI education. The research integrates the Technology Acceptance Model, AI literacy framework, and sustainable development theory to explore the relationships among users' perceptions, applications, trust in AI, and ESD, thereby enhancing explanatory power. Path analysis is employed to identify and quantify the influence of each factor on ESD. This study offers practical recommendations for the digital transformation of university libraries, providing actionable guidance for optimizing digital services and leveraging Large-Scale Artificial Intelligence Models to deepen ESD practices and enhance educational empowerment.\u003c/p\u003e \u003cp\u003eThe structure of this study is organized as follows: following a literature review, hypotheses are proposed based on the Technology Acceptance Model, AI literacy framework, and sustainable development theory. Subsequently, the study elaborates on sample selection, data sources, variable definitions, and model construction. The regression results of the model are then analyzed, followed by robustness and heterogeneity tests of the applied models. Finally, the study concludes with a summary of findings and offers policy recommendations.\u003c/p\u003e"},{"header":"2. Related work","content":"\u003cdiv id=\"Sec3\" class=\"Section2\"\u003e \u003ch2\u003e2.1. Utilization of Large-Scale Artificial Intelligence Models in Library Information Retrieval Systems\u003c/h2\u003e \u003cp\u003eCurrently, the rapid advancement of Large-Scale Models is driving a new wave of innovation in artificial intelligence. To advance the digital transformation of libraries, the Chinese government has introduced a series of policies to strengthen the strategic development of smart libraries. Within the framework of 'Large-Scale Models\u0026thinsp;+\u0026thinsp;Smart Libraries', Large-Scale Artificial Intelligence Models serve as the core for extending smart library service applications. These services provide readers with a unified access point, offering diverse functionalities such as consultation and question-answering, search recommendations, and innovative reading experiences [\u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e8\u003c/span\u003e]. As key hubs for knowledge innovation and talent cultivation, university libraries play a critical role in supporting academic research, serving the learning needs of students and faculty, and advancing education. The deployment of Large-Scale Artificial Intelligence Models in information retrieval systems has become an inevitable choice for these libraries to align with contemporary technological trends. Currently, university libraries are actively implementing localized Large-Scale Artificial Intelligence Models as a strategic measure to deepen digital transformation and innovate service models. According to the 2024 annual report of the International Federation of Library Associations and Institutions (IFLA), over 78% of libraries worldwide have initiated digital transformation programs, with Chinese public libraries leading globally, achieving a digital penetration rate growing at an annual average of 15% [\u003cspan citationid=\"CR9\" class=\"CitationRef\"\u003e9\u003c/span\u003e].\u003c/p\u003e \u003cp\u003eIn 1986, Davis introduced the Technology Acceptance Model, which posits that perceived usefulness and perceived ease of use are the primary determinants of user acceptance when encountering new technology [\u003cspan citationid=\"CR10\" class=\"CitationRef\"\u003e10\u003c/span\u003e]. Perceived usefulness refers to the extent to which users believe the technology enhances their work or learning efficiency, while perceived ease of use reflects the level of convenience users experience when interacting with the system. Together, these factors drive users' technology acceptance behavior.\u003c/p\u003e \u003cp\u003eThe application of Large-Scale Artificial Intelligence Models to library information retrieval systems enables intelligent filtering and recommendation functions, assisting users in quickly accessing high-quality information and boosting their confidence in obtaining reliable resources. Users with strong information assessment capability can more effectively assess the accuracy and value of system outputs, allowing them to explore the system's potential functionalities more deeply. This enhances their perception of the system's usefulness. The system's efficient feedback further encourages these users to continuously improve their information verification and critical analysis skills during sustained use, creating a positive feedback loop of 'system use\u0026mdash;strengthened information assessment capability\u0026mdash;increased perceived usefulness\u0026mdash;higher usage frequency'. Based on this, Hypothesis 1a is proposed: The use of AI-powered library information retrieval systems can enhance users' information assessment capabilities.\u003c/p\u003e \u003cp\u003eFrom the perspective of perceived ease of use, AI-powered library information retrieval systems feature user-friendly interface designs and diverse intelligent services, significantly lowering the operational barriers for users and enhancing their perception of the system's ease of use. Users with higher digital literacy can more quickly master the system's complex interaction logic, enabling them to utilize its functions more efficiently for retrieval and other tasks. This positive user experience further increases their frequency of system use [\u003cspan citationid=\"CR11\" class=\"CitationRef\"\u003e11\u003c/span\u003e]. The process of frequent system interaction inherently involves users accumulating digital skills through practice, including data processing, tool operation, and algorithm understanding and application, thereby contributing to an overall improvement in their digital literacy [\u003cspan citationid=\"CR12\" class=\"CitationRef\"\u003e12\u003c/span\u003e]. Accordingly, Hypothesis 1b is proposed: The use of AI-powered library information retrieval systems can enhance users' digital literacy.\u003c/p\u003e \u003cp\u003eIn 2021, Davy Ng and colleagues proposed an AI literacy framework encompassing four dimensions: 'Understanding AI', 'Applying AI', 'Evaluating and Creating AI', and 'AI Ethics' [\u003cspan citationid=\"CR13\" class=\"CitationRef\"\u003e13\u003c/span\u003e]. This framework is well-suited for examining how the application of Large-Scale Artificial Intelligence Models to library information retrieval systems can enhance ESD. Within the 'Understanding AI' dimension, university libraries deploying localized Large-Scale Artificial Intelligence Models provide users with a foundation for comprehending the underlying logic of AI technologies. By leveraging information retrieval systems powered by Large-Scale Artificial Intelligence Models, libraries educate users about the data processing mechanisms of these models, strengthening their foundational understanding of privacy and security awareness and establishing a cognitive basis for privacy protection. During the application phase of AI-powered library information retrieval systems, the system processes search commands and identifies personal preferences based on user inputs. Users with heightened privacy and security awareness are more likely to scrutinize the compliance of data collection, storage, and usage, adopting standardized operational behaviors to mitigate data breach risks. They also provide ongoing feedback regarding data security, model safety, and privacy protection needs, prompting libraries to enhance data governance and reinforcing users' awareness of digital ecosystem sustainability. In the 'Evaluating and Creating AI' phase, users' privacy and security awareness is further deepened. When assessing the quality, compliance, and safety of search results generated by Large-Scale Artificial Intelligence Models, privacy and security become critical evaluation metrics. Upon identifying potential data infringements or privacy risks in search outputs, users are prompted to critically examine the system's data ethics, continuously broadening and deepening their privacy and security awareness [\u003cspan citationid=\"CR14\" class=\"CitationRef\"\u003e14\u003c/span\u003e]. By leveraging the deployment and application of Large-Scale Artificial Intelligence Models, universities integrate privacy protection and data security into technical ethics education programs. Through targeted lectures, technical training, and classroom instruction, knowledge about data sovereignty and privacy is imparted to users, simultaneously enhancing their privacy and security awareness and advancing ESD. Based on this, Hypothesis 1c is proposed: The use of AI-powered library information retrieval systems can enhance users' privacy and security awareness.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec4\" class=\"Section2\"\u003e \u003ch2\u003e2.2. Education for Sustainable Development\u003c/h2\u003e \u003cp\u003eThe Sustainable Development Goals (SDGs) 4.7 represent a core component of the United Nations sustainable development agenda, with a particular focus on the quality and purpose of education. It's closely intertwined with ESD, drawing its vision from the United Nations Decade of ESD (2005\u0026ndash;2014) and the subsequent Global Action Programme (2015\u0026ndash;2019). Specifically, SDG 4.7 aims to ensure that by 2030, all learners acquire the knowledge and skills necessary for sustainable development. This is achieved through education on sustainable development, sustainable lifestyles, human rights, and gender equality, while promoting a culture of peace and non-violence, fostering global citizenship, and recognizing cultural diversity and its contributions to sustainable development [\u003cspan citationid=\"CR15\" class=\"CitationRef\"\u003e15\u003c/span\u003e]. ESD embodies an ethic of 'care' and 'concern', aimed at equipping learners with the knowledge, skills, and mindset to act for the planet by connecting their minds, hands, and hearts. It reorients education to enable learners of all ages to take action for environmental integrity, economic viability, cultural diversity, and a more equitable society. ESD emphasizes interdisciplinarity, fostering systems thinking to address the interconnected complexities of contemporary global challenges.\u003c/p\u003e \u003cp\u003eFrom the perspective of the Technology Acceptance Model, users' acceptance of Large-Scale Artificial Intelligence Models in library information retrieval systems hinges on their perceptions of the technology's usefulness and ease of use. Users with strong information assessment capability can assess and filter search results, identifying valuable content, which encourages more active engagement with the system for learning and enhances their ESD. Users with high digital literacy are adept at operating information retrieval systems, fully leveraging the capabilities of Large-Scale Artificial Intelligence Models to efficiently access needed information. This positive experience of ease of use promotes frequent system use, increasing exposure to ESD content and ultimately strengthening its educational impact.\u003c/p\u003e \u003cp\u003eThe AI literacy framework emphasizes the knowledge, skills, and attitudes that individuals should possess in AI-driven environments. Users with privacy and security awareness understand concepts such as data security and information protection, enabling them to identify potential risks in acquiring and using information on ESD. This awareness helps prevent disruptions to educational activities caused by data breaches or misuse, ensuring the continuity and effectiveness of learning. Additionally, this awareness encourages users to acquire secure operational skills, such as encryption and authentication, ensuring that data and personal privacy are adequately protected when utilizing Large-Scale Artificial Intelligence Models in library information retrieval systems to access educational resources for sustainable development. This creates a secure environment conducive to in-depth learning [\u003cspan citationid=\"CR16\" class=\"CitationRef\"\u003e16\u003c/span\u003e]. From an attitudinal perspective, privacy and security awareness fosters a sense of responsibility among users to uphold information ethics and maintain a healthy digital ecosystem. Consequently, in engaging with ESD, users not only focus on their own knowledge acquisition but also actively contribute to building a fair, healthy, and orderly digital education environment, thereby supporting the realization of educational goals for sustainable development [\u003cspan citationid=\"CR16\" class=\"CitationRef\"\u003e16\u003c/span\u003e].\u003c/p\u003e \u003cp\u003eBased on the foregoing analysis, the following hypotheses are proposed:\u003c/p\u003e \u003cp\u003e \u003cstrong\u003eHypothesis\u003c/strong\u003e \u003cp\u003ea: Users' information assessment capability is positively correlated with ESD.\u003c/p\u003e \u003c/p\u003e \u003cp\u003e \u003cstrong\u003eHypothesis\u003c/strong\u003e \u003cp\u003eb: Users' digital literacy is positively correlated with ESD.\u003c/p\u003e \u003c/p\u003e \u003cp\u003e \u003cstrong\u003eHypothesis\u003c/strong\u003e \u003cp\u003ec: Users' privacy and security awareness is positively correlated with ESD.\u003c/p\u003e \u003c/p\u003e \u003cp\u003e2.3. Application of Large-Scale Artificial Intelligence Models in Library Information Retrieval Systems and Education for Sustainable Development\u003c/p\u003e\u003cp\u003eCurrent academic research on ESD primarily focuses on curriculum design, pedagogical approaches, technology integration, student engagement, and evaluation and policy support [\u003cspan additionalcitationids=\"CR18 CR19 CR20\" citationid=\"CR17\" class=\"CitationRef\"\u003e17\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR21\" class=\"CitationRef\"\u003e21\u003c/span\u003e]. Literature on the application of technology in ESD aligns with theoretical frameworks and provides a robust foundation for this study. Applications of artificial intelligence in academia and education encompass educational support, constructive feedback, assessment, customized curricula, personalized career guidance, and mental health support [\u003cspan citationid=\"CR19\" class=\"CitationRef\"\u003e19\u003c/span\u003e]. Research indicates that Large-Scale Artificial Intelligence Models can significantly enhance valuable learning outcomes for higher education students [\u003cspan citationid=\"CR22\" class=\"CitationRef\"\u003e22\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR23\" class=\"CitationRef\"\u003e23\u003c/span\u003e]. This offers concrete practical references and theoretical grounding for advancing ESD through the integration of Large-Scale Artificial Intelligence Models.\u003c/p\u003e \u003cp\u003eAlthough most studies support the notion that Large-Scale Artificial Intelligence Models enhance information access and learning efficiency, thereby promoting ESD, some controversies persist. Contrasting research highlights that ethical issues in AI technologies may undermine educational outcomes. Appleton argues that AI systems must address biases and privacy concerns to ensure their legitimate and compliant application in education, as failure to do so could negatively impact the learning environment of ESD [\u003cspan citationid=\"CR24\" class=\"CitationRef\"\u003e24\u003c/span\u003e]. Similarly, Fu emphasizes that excessive reliance on AI by users may diminish critical thinking skills, posing a challenge to the systemic objectives of ESD [\u003cspan citationid=\"CR25\" class=\"CitationRef\"\u003e25\u003c/span\u003e].\u003c/p\u003e \u003cp\u003eBased on the preceding analysis, the following hypotheses are proposed:\u003c/p\u003e \u003cp\u003e \u003cstrong\u003eHypothesis\u003c/strong\u003e \u003cp\u003ea: The use of AI-powered library information retrieval systems is positively correlated with ESD.\u003c/p\u003e \u003c/p\u003e \u003cp\u003eSustainable development theory emphasizes the harmonious advancement of economic, social, and environmental dimensions, with education serving as a critical pathway to achieving these goals. Within library information retrieval systems, users' information assessment capability enables them to filter high-quality educational resources aligned with sustainable development principles, fostering greater awareness of sustainability issues and cultivating appropriate values and a sense of responsibility. Digital literacy empowers users to leverage Large-Scale Artificial Intelligence Models to bridge information gaps, accessing educational resources for sustainable development and promoting equitable sharing, thereby enhancing the inclusivity of ESD. Privacy and security awareness fosters a healthy digital ecosystem for improving ESD, ensuring the lawful and ethical use of information and facilitating the orderly conduct of educational activities. Together, these factors synergistically advance the objectives of ESD. Similarly, Lo underscores the need for libraries deploying Large-Scale Artificial Intelligence Models to simultaneously enhance the AI literacy of users and other stakeholders [\u003cspan citationid=\"CR26\" class=\"CitationRef\"\u003e26\u003c/span\u003e]. This involves engaging librarians in the implementation of AI, developing AI usage guidelines, and advocating that academic libraries proactively embrace AI to ensure its effective, ethical, and responsible applications in library services and operations. Based on this, the following Hypothesis is proposed:\u003c/p\u003e \u003cp\u003e \u003cstrong\u003eHypothesis\u003c/strong\u003e \u003cp\u003eb: Users' information assessment capability, digital literacy, and privacy and security awareness partially moderate the relationship posited in Hypothesis 3a.\u003c/p\u003e \u003c/p\u003e \u003c/div\u003e"},{"header":"3. Materials and Methods","content":"\u003cdiv id=\"Sec6\" class=\"Section2\"\u003e \u003ch2\u003e3.1. Sample selection and data sources\u003c/h2\u003e \u003cp\u003eThis study collects data through a survey questionnaire. Informed consent was obtained from all participants and / or their legal guardians. This study complies with ethical standards and is approved by the Ethics Committee of School of Marxism, Harbin University of Science and Technology. Participants' written informed consents are obtained. All experiments were performed in accordance with relevant guidelines and regulations.\u003c/p\u003e \u003cp\u003eIn May 2025, a pre-test and pilot study was conducted with 48 students to assess the validity and reliability of the questionnaire. This preliminary survey helped evaluate the relevance, clarity, and comprehensiveness of the included scales, ensuring their suitability for the target population. In June 2025, a formal survey was administered to 315 respondents via the Wenjuanxing platform. The purpose of the questionnaire was to evaluate the pedagogical implications of applying Large-Scale Artificial Intelligence Models to library information retrieval systems, grounded in the theoretical framework of AI literacy. The survey was structured around the AI literacy framework and the United Nations Sustainable Development Goal 4 (SDG 4) framework, encompassing eight categories: understanding and comprehending AI; using and applying AI; evaluating and creating AI; AI ethics; digital literacy; cybersecurity awareness; educational equity and learning outcomes; and educational opportunities, skills, and sustainable development. No personal information was collected during the survey, ensuring complete anonymity. The collected data were securely stored and kept confidential to mitigate any risks to respondents' interests.\u003c/p\u003e \u003cp\u003eParticipants included a small number of high school students, as well as undergraduate and postgraduate students. Accordingly, the study employed a non-probability purposive sampling method to investigate the experiences of higher education students using Large-Scale Artificial Intelligence Models in library information retrieval systems. Etikan notes that non-probability purposive sampling is one of the most common techniques for selecting respondents with knowledge or experience relevant to a specific research question [\u003cspan citationid=\"CR27\" class=\"CitationRef\"\u003e27\u003c/span\u003e]. Therefore, selecting university students pursuing bachelor's, master's, and doctoral degrees was appropriate, as they represent a significant segment of higher education and have either experienced or are currently engaging with digital learning environments facilitated by Large-Scale Artificial Intelligence Models in library information retrieval systems. This non-probability sampling technique enabled researchers to deliberately select respondents with specific characteristics, ensuring the integrity and relevance of the collected data [\u003cspan citationid=\"CR28\" class=\"CitationRef\"\u003e28\u003c/span\u003e].\u003c/p\u003e \u003cp\u003eData processing and analysis were conducted using SPSS, with subsets of the original dataset extracted based on specific criteria for further analysis. After excluding data from high school students and invalid responses, 205 valid responses remained for analysis.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec7\" class=\"Section2\"\u003e \u003ch2\u003e3.2. Variable Definitions\u003c/h2\u003e \u003cdiv id=\"Sec8\" class=\"Section3\"\u003e \u003ch2\u003e3.2.1. Dependent variable\u003c/h2\u003e \u003cp\u003eThe dependent variable is ESD. It is measured by averaging the scores of items in the sixth section (Educational Equity and Learning Outcomes) and the seventh section (Educational Opportunities, Skills, and Sustainable Development) of the questionnaire. Higher values indicate a greater level of ESD.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec9\" class=\"Section3\"\u003e \u003ch2\u003e3.2.2 Independent variable\u003c/h2\u003e \u003cp\u003eThe independent variable is the ability to use library information retrieval systems, assessed through AI usage capability. This is calculated by averaging the scores of all items in the second section (Using and Applying AI) of the questionnaire. Higher values reflect stronger proficiency in using library information retrieval systems.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec10\" class=\"Section3\"\u003e \u003ch2\u003e3.2.3 Mediating variables\u003c/h2\u003e \u003cp\u003eThis study selects digital literacy, information assessment capability, and privacy and security awareness as mediating variables to further analyze the pathways through which the application of Large-Scale Artificial Intelligence Models in library information retrieval systems impacts ESD. Digital literacy and information assessment capability are measured by averaging the scores of items in the fourth section (Digital Literacy) and the third section (AI Ethics and Assessment) of the questionnaire, respectively. Privacy and security awareness is assessed based on responses to a single questionnaire item, specifically the level of agreement with the statement, 'I believe it is very necessary to protect personal data in library AI systems' (rated on a scale from 1\u0026thinsp;=\u0026thinsp;Strongly Disagree to 5\u0026thinsp;=\u0026thinsp;Strongly Agree). The resulting value reflects the individual's emphasis on data privacy.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec11\" class=\"Section3\"\u003e \u003ch2\u003e3.2.4 Control variables\u003c/h2\u003e \u003cp\u003eDrawing on existing literature, this study incorporates the following control variables to ensure the robustness of the findings:\u003c/p\u003e \u003cp\u003e \u003col\u003e \u003cspan\u003e \u003cli\u003e \u003cp\u003eGender: This refers to the biological sex of the participants and is used to explore potential differences in behavioral preferences and acceptance levels across genders. Variations in information retrieval habits and learning styles between genders may influence the accuracy of the study's results.\u003c/p\u003e \u003c/li\u003e \u003c/span\u003e \u003cspan\u003e \u003cli\u003e \u003cp\u003eAge: This represents the age group of the participants and is included to assess whether individuals in different age brackets exhibit varying responses when using library information retrieval systems embedded with Large-Scale Artificial Intelligence Models [\u003cspan citationid=\"CR29\" class=\"CitationRef\"\u003e29\u003c/span\u003e]. Age differences are often associated with disparities in learning capabilities and technology acceptance, which can significantly impact the study's conclusions.\u003c/p\u003e \u003c/li\u003e \u003c/span\u003e \u003cspan\u003e \u003cli\u003e \u003cp\u003eGrades: This indicates the educational stage of the students (e.g., undergraduate\u0026rsquo;s, master's, or doctoral level) and is used to analyze whether differences in knowledge reserves and learning objectives among students at various grades lead to differing outcomes when utilizing library information retrieval systems.\u003c/p\u003e \u003c/li\u003e \u003c/span\u003e \u003cspan\u003e \u003cli\u003e \u003cp\u003eSubject Background: This refers to the participants' academic fields [\u003cspan citationid=\"CR30\" class=\"CitationRef\"\u003e30\u003c/span\u003e]. It is included to investigate whether groups with different academic backgrounds produce varied results when accessing ESD resources through library information retrieval systems, due to differences in professional thinking and information needs.\u003c/p\u003e \u003c/li\u003e \u003c/span\u003e \u003cspan\u003e \u003cli\u003e \u003cp\u003eAI Education Level: This measures the extent to which participants have received education related to artificial intelligence [\u003cspan citationid=\"CR31\" class=\"CitationRef\"\u003e31\u003c/span\u003e]. It is used to evaluate whether individuals with varying levels of AI education exhibit different outcomes when using library information retrieval systems, influenced by their familiarity with AI technologies and application proficiency. AI education level significantly affects respondents' operational proficiency and system utilization efficiency, making it a critical factor.\u003c/p\u003e \u003c/li\u003e \u003c/span\u003e \u003c/ol\u003e \u003c/p\u003e \u003cp\u003eTable\u0026nbsp;\u003cspan refid=\"Tab1\" class=\"InternalRef\"\u003e1\u003c/span\u003e presents the definitions of the main variables in the research analysis.\u003c/p\u003e \u003cp\u003e \u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"Yes\" id=\"Tab1\" border=\"1\"\u003e \u003ccaption language=\"En\"\u003e \u003cdiv class=\"CaptionNumber\"\u003eTable 1\u003c/div\u003e \u003cdiv class=\"CaptionContent\"\u003e \u003cp\u003eDefinitions and measurement methods of key variables.\u003c/p\u003e \u003c/div\u003e \u003c/caption\u003e \u003ccolgroup cols=\"4\"\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c3\" colnum=\"3\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c4\" colnum=\"4\"\u003e\u003c/div\u003e \u003cthead\u003e \u003ctr\u003e \u003cth align=\"left\" colname=\"c1\"\u003e \u003cp\u003eVariable Type\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c2\"\u003e \u003cp\u003eVariable Name\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c3\"\u003e \u003cp\u003enotation\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c4\"\u003e \u003cp\u003eDefinition and Measurement\u003c/p\u003e \u003c/th\u003e \u003c/tr\u003e \u003c/thead\u003e \u003ctbody\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eDependent Variable\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eEducation for Sustainable Development\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e\u003cem\u003esustEdu\u003c/em\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003eCalculate the average scores of all items in Section 6 (Educational Equity and Learning Outcomes) and Section 7 (Educational Opportunities, Skills, and Sustainable Development) of the questionnaire.\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eIndependent Variable\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eAI Usage capability\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e\u003cem\u003eaiSkill\u003c/em\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003eCalculate the average scores of all items in the dimension of Section 2 (Use and Application of AI) of the questionnaire.\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\" morerows=\"2\" rowspan=\"3\"\u003e \u003cp\u003eMediating Variables\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eDigital Literacy\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e\u003cem\u003edigLit\u003c/em\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003eCalculate the average scores of all items in the dimension of Section 4 (Digital Literacy) of the questionnaire.\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eInformation Assessment Capability\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e\u003cem\u003einfoAssess\u003c/em\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003eCalculate the average scores of all items in the dimension of Section 3 (AI Ethics and Assessment) of the questionnaire.\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003ePrivacy and Security Awareness\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e\u003cem\u003epvtSec\u003c/em\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003eMeasure the degree of agreement with the statement 'I believe it is very necessary to protect personal data in library AI systems' (1\u0026thinsp;=\u0026thinsp;strongly disagree, 5\u0026thinsp;=\u0026thinsp;strongly agree) among respondents, using it as a proxy variable for individuals' emphasis on data privacy.\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\" morerows=\"4\" rowspan=\"5\"\u003e \u003cp\u003eControl Variables\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eGender\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e\u003cem\u003eGender\u003c/em\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e'Male' = 0; 'Female' = 1.\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eGrade\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e\u003cem\u003eGrade\u003c/em\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e'Freshman' = 1; 'Sophomore' = 2; 'Junior' = 3; 'Senior' = 4; 'Graduate (Master's/PhD)' = 5.\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eAI Education Level\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e\u003cem\u003eaiEduLevel\u003c/em\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e'No exposure' = 1; 'Basic concept learning (less than 10 hours)' = 2; 'Systematic course learning (10\u0026ndash;30 hours)' = 3; 'Professional application training (more than 30 hours)' = 4.\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eProvince\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e\u003cem\u003eprov\u003c/em\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003eClassify Chinese provinces into 'developed regions' and 'developing regions' based on their level of economic development, assigning a value of 1 to 'developed regions' and a value of 2 to 'developing regions'.\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eSubject Background\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e\u003cem\u003esubjBkg\u003c/em\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e'Humanities and Social Sciences' = 1; 'Natural Sciences' = 2.\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003c/tbody\u003e \u003c/colgroup\u003e \u003c/table\u003e\u003c/div\u003e \u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec12\" class=\"Section3\"\u003e \u003ch2\u003e3.2.5. Model construction\u003c/h2\u003e \u003cp\u003eTo analyze the impact of applying Large-Scale Artificial Intelligence Models to library information retrieval systems on ESD, Model (1) is constructed as follows:\u003c/p\u003e \u003cp\u003e \u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"No\" id=\"Taba\" border=\"1\"\u003e \u003ccolgroup cols=\"2\"\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e \u003ctbody\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u003cem\u003esustEdui\u0026thinsp;=\u0026thinsp;β\u003c/em\u003e\u003csub\u003e\u003cem\u003e0\u003c/em\u003e\u003c/sub\u003e\u0026thinsp;\u003cem\u003e+\u0026thinsp;β\u003c/em\u003e\u003csub\u003e\u003cem\u003e1\u003c/em\u003e\u003c/sub\u003e\u003cem\u003e∙aiSkilli\u0026thinsp;+\u0026thinsp;β\u003c/em\u003e\u003csub\u003e\u003cem\u003e2\u003c/em\u003e\u003c/sub\u003e\u003cem\u003e∙Genderi\u0026thinsp;+\u0026thinsp;β\u003c/em\u003e\u003csub\u003e\u003cem\u003e3\u003c/em\u003e\u003c/sub\u003e\u003cem\u003e∙Gradei\u0026thinsp;+\u0026thinsp;β\u003c/em\u003e\u003csub\u003e\u003cem\u003e4\u003c/em\u003e\u003c/sub\u003e\u003cem\u003e∙aiEduLeveli\u0026thinsp;+\u0026thinsp;β\u003c/em\u003e\u003csub\u003e\u003cem\u003e5\u003c/em\u003e\u003c/sub\u003e\u003cem\u003e∙provi\u003c/em\u003e\u003c/p\u003e \u003cp\u003e\u003cem\u003e+β\u003c/em\u003e\u003csub\u003e\u003cem\u003e6\u003c/em\u003e\u003c/sub\u003e\u003cem\u003e∙subjBkgi+ϵ\u003c/em\u003e\u003csub\u003e\u003cem\u003ei\u003c/em\u003e\u003c/sub\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e(1)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003c/tbody\u003e \u003c/colgroup\u003e \u003c/table\u003e\u003c/div\u003e \u003c/p\u003e\u003cp\u003eTo further investigate the mechanistic roles of digital literacy, information assessment capability, and privacy and security awareness, Model (2) is constructed as follows:\u003c/p\u003e \u003cp\u003e \u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"No\" id=\"Tabb\" border=\"1\"\u003e \u003ccolgroup cols=\"2\"\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e \u003ctbody\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u003cem\u003esustEdui\u0026thinsp;=\u0026thinsp;γ\u003c/em\u003e\u003csub\u003e\u003cem\u003e0\u003c/em\u003e\u003c/sub\u003e\u0026thinsp;\u003cem\u003e+\u0026thinsp;γ\u003c/em\u003e\u003csub\u003e\u003cem\u003e1\u003c/em\u003e\u003c/sub\u003e\u003cem\u003e∙aiSkilli\u0026thinsp;+\u0026thinsp;γ\u003c/em\u003e\u003csub\u003e\u003cem\u003e2\u003c/em\u003e\u003c/sub\u003e\u003cem\u003e∙dgLiti\u0026thinsp;+\u0026thinsp;γ\u003c/em\u003e\u003csub\u003e\u003cem\u003e3\u003c/em\u003e\u003c/sub\u003e\u003cem\u003e∙infoAssessi\u0026thinsp;+\u0026thinsp;γ\u003c/em\u003e\u003csub\u003e\u003cem\u003e4\u003c/em\u003e\u003c/sub\u003e\u003cem\u003e∙pvtSeci+Controls\u0026thinsp;+\u0026thinsp;u\u003c/em\u003e\u003csub\u003e\u003cem\u003ei\u003c/em\u003e\u003c/sub\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e(2)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003c/tbody\u003e \u003c/colgroup\u003e \u003c/table\u003e\u003c/div\u003e \u003c/p\u003e \u003cp\u003eIn Models 1 and 2, \u003cem\u003esustEdu\u003c/em\u003e represents ESD and serves as the dependent variable. The independent variable \u003cem\u003eaiSkill\u003c/em\u003e denotes the respondents' proficiency in using library AI information retrieval systems. The mediating variables \u003cem\u003edigLit\u003c/em\u003e, \u003cem\u003einfoAssess\u003c/em\u003e, and \u003cem\u003epvtSec\u003c/em\u003e represent digital literacy, information assessment capability, and privacy and security awareness, respectively. The control variables \u003cem\u003eGender\u003c/em\u003e, \u003cem\u003eGrade\u003c/em\u003e, \u003cem\u003eaiEduLevel\u003c/em\u003e, \u003cem\u003eprov\u003c/em\u003e, and \u003cem\u003esubjBkg\u003c/em\u003e correspond to gender, academic year, AI education level, province, and subject background, respectively, and are collectively denoted as Controls in Model 2. The subscript \u003cem\u003ei\u003c/em\u003e indicates individual respondents. The terms \u003cem\u003eϵ\u003c/em\u003e\u003csub\u003e\u003cem\u003ei\u003c/em\u003e\u003c/sub\u003e and \u003cem\u003eu\u003c/em\u003e\u003csub\u003e\u003cem\u003ei\u003c/em\u003e\u003c/sub\u003e represent the random error terms for Model 1 and Model 2, respectively.\u003c/p\u003e \u003c/div\u003e \u003c/div\u003e \u003cdiv id=\"Sec13\" class=\"Section2\"\u003e \u003ch2\u003e3.3. Common Method Variance Analysis\u003c/h2\u003e \u003cp\u003eData were collected through a survey questionnaire, with respondents primarily completing the questionnaire based on their subjective perceptions, which may introduce common method variance. To address this, we first employed Harman's single-factor test to assess the extent of common method variance in the data after collection. The results indicated that the maximum factor variance explained was 27.192%, which is below the critical threshold of 40%. This suggests that the issue of common method variance in the data was effectively controlled.\u003c/p\u003e \u003c/div\u003e"},{"header":"4. Results and Discussion","content":"\u003cdiv id=\"Sec15\" class=\"Section2\"\u003e \u003ch2\u003e4.1. Demographics statistics\u003c/h2\u003e \u003cp\u003e \u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"Yes\" id=\"Tab2\" border=\"1\"\u003e \u003ccaption language=\"En\"\u003e \u003cdiv class=\"CaptionNumber\"\u003eTable 2\u003c/div\u003e \u003cdiv class=\"CaptionContent\"\u003e \u003cp\u003eDemographics statistics.\u003c/p\u003e \u003c/div\u003e \u003c/caption\u003e \u003ccolgroup cols=\"4\"\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c3\" colnum=\"3\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c4\" colnum=\"4\"\u003e\u003c/div\u003e \u003cthead\u003e \u003ctr\u003e \u003cth align=\"left\" colname=\"c1\"\u003e \u003cp\u003eVariable\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c2\"\u003e \u003cp\u003eType\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c3\"\u003e \u003cp\u003eFrequency\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c4\"\u003e \u003cp\u003ePercent\u003c/p\u003e \u003c/th\u003e \u003c/tr\u003e \u003c/thead\u003e \u003ctbody\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\" morerows=\"1\" rowspan=\"2\"\u003e \u003cp\u003e\u003cem\u003eGender\u003c/em\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eFemale\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e111\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e54.1\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eMale\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e94\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e45.9\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\" morerows=\"4\" rowspan=\"5\"\u003e \u003cp\u003e\u003cem\u003eGrade\u003c/em\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eFreshman\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e36\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e17.6\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eSophomore\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e54\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e26.3\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eJunior\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e57\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e27.8\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eSenior\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e46\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e22.4\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eGraduate (Master/PhD)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e12\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e5.9\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\" morerows=\"3\" rowspan=\"4\"\u003e \u003cp\u003e\u003cem\u003eaiEduLevel\u003c/em\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eNever touched\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e23\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e11.2\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eBasic concept learning (less than 10 class hours)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e125\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e61\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eSystematic course learning (10\u0026ndash;30 class hours)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e53\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e25.9\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eProfessional application training (more than 30 class hours)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e4\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e2\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\" morerows=\"1\" rowspan=\"2\"\u003e \u003cp\u003e\u003cem\u003eprov\u003c/em\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003edeveloped regions\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e84\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e41\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003edeveloping regions\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e121\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e59\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\" morerows=\"1\" rowspan=\"2\"\u003e \u003cp\u003e\u003cem\u003esubjBkg\u003c/em\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eHumanities and Social Sciences\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e114\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e55.6\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eNatural Sciences\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e91\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e44.4\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003c/tbody\u003e \u003c/colgroup\u003e \u003c/table\u003e\u003c/div\u003e \u003c/p\u003e \u003cp\u003eA total of 315 respondents participated in our questionnaire survey, with 205 valid samples remaining after data cleaning. Table\u0026nbsp;\u003cspan refid=\"Tab2\" class=\"InternalRef\"\u003e2\u003c/span\u003e presents the demographic results. Of the respondents, 111 were female, and 94 were male. The majority were undergraduate students, accounting for 94.1% of the sample. Additionally, 61% of respondents had received formal education on basic concepts related to artificial intelligence. Regarding the respondents' provinces and subject backgrounds, the distributions were relatively balanced. Specifically, 41% of respondents were from economically developed regions, and 55.6% had a subject background in the humanities and social sciences.\u003c/p\u003e \u003cp\u003e \u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"Yes\" id=\"Tab3\" border=\"1\"\u003e \u003ccaption language=\"En\"\u003e \u003cdiv class=\"CaptionNumber\"\u003eTable 3\u003c/div\u003e \u003cdiv class=\"CaptionContent\"\u003e \u003cp\u003eCorrelation Analysis.\u003c/p\u003e \u003c/div\u003e \u003c/caption\u003e \u003ccolgroup cols=\"11\"\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c3\" colnum=\"3\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c4\" colnum=\"4\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c5\" colnum=\"5\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c6\" colnum=\"6\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c7\" colnum=\"7\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c8\" colnum=\"8\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c9\" colnum=\"9\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c10\" colnum=\"10\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c11\" colnum=\"11\"\u003e\u003c/div\u003e \u003cthead\u003e \u003ctr\u003e \u003cth align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/th\u003e \u003cth align=\"left\" colname=\"c2\"\u003e \u003cp\u003e\u003cem\u003esustEdu\u003c/em\u003e\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c3\"\u003e \u003cp\u003e\u003cem\u003eaiSkill\u003c/em\u003e\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c4\"\u003e \u003cp\u003e\u003cem\u003einfoAssess\u003c/em\u003e\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c5\"\u003e \u003cp\u003e\u003cem\u003edigLit\u003c/em\u003e\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c6\"\u003e \u003cp\u003e\u003cem\u003epvtSec\u003c/em\u003e\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c7\"\u003e \u003cp\u003e\u003cem\u003eprov\u003c/em\u003e\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c8\"\u003e \u003cp\u003e\u003cem\u003eGender\u003c/em\u003e\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c9\"\u003e \u003cp\u003e\u003cem\u003eGrade\u003c/em\u003e\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c10\"\u003e \u003cp\u003e\u003cem\u003esubjBkg\u003c/em\u003e\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c11\"\u003e \u003cp\u003e\u003cem\u003eaiEduLevel\u003c/em\u003e\u003c/p\u003e \u003c/th\u003e \u003c/tr\u003e \u003c/thead\u003e \u003ctbody\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u003cem\u003esustEdu\u003c/em\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e1\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c9\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c10\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c11\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u003cem\u003eaiSkill\u003c/em\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e0.688***\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e1\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c9\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c10\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c11\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u003cem\u003einfoAssess\u003c/em\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e0.561***\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e0.560***\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e1\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c9\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c10\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c11\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u003cem\u003edigLit\u003c/em\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e0.763***\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e0.826***\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e0.621***\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e1\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c9\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c10\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c11\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u003cem\u003epvtSec\u003c/em\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e0.403***\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e0.292***\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e0.256***\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e0.334***\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e1\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c9\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c10\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c11\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u003cem\u003eprov\u003c/em\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e0.105\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e0.013\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e0.062\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e0.053\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e-0.001\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e1\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c9\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c10\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c11\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u003cem\u003eGender\u003c/em\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e0.139**\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e0.098\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e-0.025\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e0.141**\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e0.113\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e-0.010\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e \u003cp\u003e1\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c9\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c10\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c11\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u003cem\u003eGrade\u003c/em\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e0.171**\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e0.173**\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e0.118*\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e0.133*\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e-0.051\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e-0.025\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e \u003cp\u003e0.011\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c9\"\u003e \u003cp\u003e1\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c10\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c11\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u003cem\u003esubjBkg\u003c/em\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e0.006\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e-0.005\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e0.035\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e0.033\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e0.018\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e0.145**\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e \u003cp\u003e-0.163**\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c9\"\u003e \u003cp\u003e0.024\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c10\"\u003e \u003cp\u003e1\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c11\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u003cem\u003eaiEduLevel\u003c/em\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e0.238***\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e0.246***\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e0.296***\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e0.306***\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e0.030\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e0.024\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e \u003cp\u003e0.022\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c9\"\u003e \u003cp\u003e0.094\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c10\"\u003e \u003cp\u003e0.109\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c11\"\u003e \u003cp\u003e1\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003c/tbody\u003e \u003c/colgroup\u003e \u003c/table\u003e\u003c/div\u003e \u003c/p\u003e \u003cp\u003eThe study employed Pearson's correlation test to examine the relationships among the variables, with the results presented in Table\u0026nbsp;\u003cspan refid=\"Tab3\" class=\"InternalRef\"\u003e3\u003c/span\u003e. The correlation coefficient between \u003cem\u003eaiSkill\u003c/em\u003e and \u003cem\u003esustEdu\u003c/em\u003e was 0.688, significant at the 1% level (denoted as ***). This indicates that the application of Large-Scale Artificial Intelligence Models in library information retrieval systems has a positive impact on ESD, providing preliminary support for Hypothesis 3a. The correlation coefficients between \u003cem\u003eaiSkill\u003c/em\u003e and \u003cem\u003einfoAssess\u003c/em\u003e, \u003cem\u003edigLit\u003c/em\u003e, and \u003cem\u003epvtSec\u003c/em\u003e were 0.560, 0.826, and 0.292, respectively, all significant at the 1% level (denoted as ***). These findings suggest positive correlations between the application of Large-Scale.\u003c/p\u003e \u003cp\u003eArtificial Intelligence Models in library information retrieval systems and information assessment capability, digital literacy, and privacy and security awareness, providing preliminary support for Hypotheses 1a, 1b, and 1c. Additionally, the correlation coefficients between \u003cem\u003einfoAssess\u003c/em\u003e, \u003cem\u003edigLit\u003c/em\u003e, \u003cem\u003epvtSec\u003c/em\u003e, and \u003cem\u003esustEdu\u003c/em\u003e were 0.561, 0.763, and 0.403, respectively, all significant at the 1% level (denoted as ***). This indicates strong positive correlations between information assessment capability, digital literacy, privacy and security awareness, and ESD, providing preliminary support for Hypotheses 2a, 2b, and 2c.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec16\" class=\"Section2\"\u003e \u003ch2\u003e4.2. Reliability and validity testing of the questionnaire\u003c/h2\u003e \u003cp\u003e \u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"Yes\" id=\"Tab4\" border=\"1\"\u003e \u003ccaption language=\"En\"\u003e \u003cdiv class=\"CaptionNumber\"\u003eTable 4\u003c/div\u003e \u003cdiv class=\"CaptionContent\"\u003e \u003cp\u003eValidity and reliability of constructs.\u003c/p\u003e \u003c/div\u003e \u003c/caption\u003e \u003ccolgroup cols=\"7\"\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c3\" colnum=\"3\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c4\" colnum=\"4\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c5\" colnum=\"5\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c6\" colnum=\"6\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c7\" colnum=\"7\"\u003e\u003c/div\u003e \u003cthead\u003e \u003ctr\u003e \u003cth align=\"left\" colname=\"c1\" morerows=\"1\" rowspan=\"2\"\u003e \u003cp\u003eVariable\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c2\" morerows=\"1\" rowspan=\"2\"\u003e \u003cp\u003eItem\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colspan=\"5\" nameend=\"c7\" namest=\"c3\"\u003e \u003cp\u003eMeasurements\u003c/p\u003e \u003c/th\u003e \u003c/tr\u003e \u003ctr\u003e \u003cth align=\"left\" colname=\"c3\"\u003e \u003cp\u003eα\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c4\"\u003e \u003cp\u003eKMO\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colspan=\"3\" nameend=\"c7\" namest=\"c5\"\u003e \u003cp\u003eBartlett\u0026rsquo;s Test of Sphericity\u003c/p\u003e \u003c/th\u003e \u003c/tr\u003e \u003c/thead\u003e \u003ctbody\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003eapproximate chi-square\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003eDegrees of Freedom\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003eSignificance\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u003cem\u003eaiSkill\u003c/em\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e6\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e0.721\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e0.800\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e203.146\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e15\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e0.000\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u003cem\u003esustEdu\u003c/em\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e12\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e0.753\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e0.818\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e341.697\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e66\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e0.000\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u003cem\u003einfoAssess\u003c/em\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e6\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e0.700\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e0.767\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e182.319\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e15\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e0.000\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u003cem\u003edigLit\u003c/em\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e6\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e0.714\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e0.803\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e181.905\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e15\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e0.000\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003c/tbody\u003e \u003c/colgroup\u003e \u003c/table\u003e\u003c/div\u003e \u003c/p\u003e \u003cp\u003eThe study utilized scales for certain variables, and thus, reliability and validity tests were conducted on these scales, with results presented in Table\u0026nbsp;\u003cspan refid=\"Tab4\" class=\"InternalRef\"\u003e4\u003c/span\u003e. Reliability Testing: The internal consistency of the scales was evaluated using Cronbach's α coefficient. The α values for all variables' scales exceeded 0.7, indicating that the measurements in this study possess high internal consistency [\u003cspan citationid=\"CR32\" class=\"CitationRef\"\u003e32\u003c/span\u003e]. Validity Testing: First, in terms of content validity, the initial measurement instrument was developed strictly based on the theoretical frameworks of artificial intelligence literacy and sustainable development. The questionnaire was refined and finalized through structured interviews with scholars in the same research field and pre-testing, ensuring strong content validity. Second, in terms of construct validity, the KMO test and Bartlett's test of sphericity were employed to assess whether the measurement items for each variable were suitable for factor analysis [\u003cspan citationid=\"CR33\" class=\"CitationRef\"\u003e33\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR34\" class=\"CitationRef\"\u003e34\u003c/span\u003e]. The KMO values for all scales exceeded the standard threshold of 0.7, and the P-values from Bartlett's test for the four variables were less than 0.001, indicating statistical significance. These results confirm the presence of common factors among the measurement item matrices, making them suitable for factor analysis and verifying construct validity.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec17\" class=\"Section2\"\u003e \u003ch2\u003e4.3 Multicollinearity analysis\u003c/h2\u003e \u003cp\u003e \u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"Yes\" id=\"Tab5\" border=\"1\"\u003e \u003ccaption language=\"En\"\u003e \u003cdiv class=\"CaptionNumber\"\u003eTable 5\u003c/div\u003e \u003cdiv class=\"CaptionContent\"\u003e \u003cp\u003eStatistical table of variance inflation factors.\u003c/p\u003e \u003c/div\u003e \u003c/caption\u003e \u003ccolgroup cols=\"3\"\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c3\" colnum=\"3\"\u003e\u003c/div\u003e \u003cthead\u003e \u003ctr\u003e \u003cth align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/th\u003e \u003cth align=\"left\" colname=\"c2\"\u003e \u003cp\u003eVIF\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c3\"\u003e \u003cp\u003e1/VIF\u003c/p\u003e \u003c/th\u003e \u003c/tr\u003e \u003c/thead\u003e \u003ctbody\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u003cem\u003edigLit\u003c/em\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e3.795\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e0.263\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u003cem\u003eaiSkill\u003c/em\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e3.251\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e0.308\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u003cem\u003einfoAssess\u003c/em\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e1.734\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e0.577\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u003cem\u003epvtSec\u003c/em\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e1.163\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e0.86\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u003cem\u003eaiEduLevel\u003c/em\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e1.151\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e0.869\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u003cem\u003eGender\u003c/em\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e1.082\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e0.924\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u003cem\u003esubjBkg\u003c/em\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e1.069\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e0.935\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u003cem\u003eGrade\u003c/em\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e1.049\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e0.953\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u003cem\u003eprov\u003c/em\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e1.03\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e0.971\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eMean VIF\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e1.703\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e.\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003c/tbody\u003e \u003c/colgroup\u003e \u003c/table\u003e\u003c/div\u003e \u003c/p\u003e \u003cp\u003eTo prevent high correlations among the explanatory and control variables, which could lead to biased model estimates and to enhance the stability and explanatory power of the model, a multicollinearity test was conducted on the independent and control variables. The results are presented in Table\u0026nbsp;\u003cspan refid=\"Tab5\" class=\"InternalRef\"\u003e5\u003c/span\u003e. Table\u0026nbsp;\u003cspan refid=\"Tab5\" class=\"InternalRef\"\u003e5\u003c/span\u003e displays the Variance Inflation Factor (VIF) and its reciprocal (1/VIF) for different variables. VIF is an indicator used in multiple regression analysis to detect multicollinearity. Multicollinearity refers to high correlations among the independent variables in a model, which can compromise the stability of the regression model and the accuracy of the coefficients of the independent variables. Typically, a VIF value exceeding 10 or 5 indicates a serious multicollinearity issue. The VIF values for all variables reported in Table\u0026nbsp;\u003cspan refid=\"Tab5\" class=\"InternalRef\"\u003e5\u003c/span\u003e are well below this threshold, suggesting that there is minimal multicollinearity among the independent and control variables. This indicates that the selection of variables is reasonable.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec18\" class=\"Section2\"\u003e \u003ch2\u003e4.4. Baseline regression result\u003c/h2\u003e \u003cp\u003e \u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"Yes\" id=\"Tab6\" border=\"1\"\u003e \u003ccaption language=\"En\"\u003e \u003cdiv class=\"CaptionNumber\"\u003eTable 6\u003c/div\u003e \u003cdiv class=\"CaptionContent\"\u003e \u003cp\u003eModel regression results.\u003c/p\u003e \u003c/div\u003e \u003c/caption\u003e \u003ccolgroup cols=\"3\"\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c3\" colnum=\"3\"\u003e\u003c/div\u003e \u003cthead\u003e \u003ctr\u003e \u003cth align=\"left\" colname=\"c1\" morerows=\"1\" rowspan=\"2\"\u003e\u0026nbsp;\u003c/th\u003e \u003cth align=\"left\" colname=\"c2\"\u003e \u003cp\u003e(1)\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c3\"\u003e \u003cp\u003e(2)\u003c/p\u003e \u003c/th\u003e \u003c/tr\u003e \u003ctr\u003e \u003cth align=\"left\" colname=\"c2\"\u003e \u003cp\u003e\u003cem\u003esustEdu\u003c/em\u003e\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c3\"\u003e \u003cp\u003e\u003cem\u003esustEdu\u003c/em\u003e\u003c/p\u003e \u003c/th\u003e \u003c/tr\u003e \u003c/thead\u003e \u003ctbody\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u003cem\u003eaiSkill\u003c/em\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e0.529***\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e0.503***\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e(13.51)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e(12.34)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u003cem\u003eprov\u003c/em\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e0.076*\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e(1.90)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u003cem\u003eGender\u003c/em\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e0.057\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e(1.44)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u003cem\u003eGrade\u003c/em\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e0.018\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e(1.03)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u003cem\u003esubjBkg\u003c/em\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e-0.001\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e(-0.03)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u003cem\u003eaiEduLevel\u003c/em\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e0.041\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e(1.30)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e_cons\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e1.997***\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e1.817***\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e(12.54)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e(10.15)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eN\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e205\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e205\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eR\u0026sup2;\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e0.474\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e0.496\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003c/tbody\u003e \u003c/colgroup\u003e \u003ctfoot\u003e \u003ctr\u003e\u003ctd colspan=\"3\"\u003e\u003cem\u003eNote: (1) *, **, and *** mean statistical significance at the 10%, 5%, and 1% level, respectively.* p\u0026thinsp;\u0026lt;\u0026thinsp;0.1, ** p\u0026thinsp;\u0026lt;\u0026thinsp;0.05, *** p\u0026thinsp;\u0026lt;\u0026thinsp;0.0; (2) t statistics in parentheses;\u003c/em\u003e\u003c/td\u003e\u003c/tr\u003e \u003c/tfoot\u003e \u003c/table\u003e\u003c/div\u003e \u003c/p\u003e \u003cp\u003eIn Table\u0026nbsp;\u003cspan refid=\"Tab6\" class=\"InternalRef\"\u003e6\u003c/span\u003e, Column (1) presents the linear regression model without control variables, while Column (2) includes control variables. The regression results confirm the Hypothesis proposed in the study, namely that the application of Large-Scale Artificial Intelligence Models in library information retrieval systems has a positive impact on ESD. In the model without control variables, the coefficient is 0.529, and after introducing control variables, the coefficient is 0.503, both significant at the 1% level, thus validating Hypothesis 3a. Additionally, the province variable, as a control variable, also has a significant effect on ESD (coefficient\u0026thinsp;=\u0026thinsp;0.076, p\u0026thinsp;\u0026lt;\u0026thinsp;0.1), with this effect being more pronounced in developing regions. Furthermore, the improved model fit underscores the importance of control variables in reducing bias and enhancing the explanatory power of the model.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec19\" class=\"Section2\"\u003e \u003ch2\u003e4.5. Mediation effect analysis\u003c/h2\u003e \u003cp\u003e \u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"Yes\" id=\"Tab7\" border=\"1\"\u003e \u003ccaption language=\"En\"\u003e \u003cdiv class=\"CaptionNumber\"\u003eTable 7\u003c/div\u003e \u003cdiv class=\"CaptionContent\"\u003e \u003cp\u003eModel regression result (\u003cem\u003einfoAssess\u003c/em\u003e).\u003c/p\u003e \u003c/div\u003e \u003c/caption\u003e \u003ccolgroup cols=\"4\"\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c3\" colnum=\"3\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c4\" colnum=\"4\"\u003e\u003c/div\u003e \u003cthead\u003e \u003ctr\u003e \u003cth align=\"left\" colname=\"c1\" morerows=\"1\" rowspan=\"2\"\u003e\u0026nbsp;\u003c/th\u003e \u003cth align=\"left\" colname=\"c2\"\u003e \u003cp\u003e(1)\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c3\"\u003e \u003cp\u003e(2)\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c4\"\u003e \u003cp\u003e(3)\u003c/p\u003e \u003c/th\u003e \u003c/tr\u003e \u003ctr\u003e \u003cth align=\"left\" colname=\"c2\"\u003e \u003cp\u003e\u003cem\u003esustEdu\u003c/em\u003e\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c3\"\u003e \u003cp\u003e\u003cem\u003einfoAssess\u003c/em\u003e\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c4\"\u003e \u003cp\u003e\u003cem\u003esustEdu\u003c/em\u003e\u003c/p\u003e \u003c/th\u003e \u003c/tr\u003e \u003c/thead\u003e \u003ctbody\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u003cem\u003eaiSkill\u003c/em\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e0.515***\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e0.592***\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e0.403***\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e(12.98)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e(9.47)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e(8.81)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u003cem\u003eprov\u003c/em\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e0.076*\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e0.056\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e0.066*\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e(1.90)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e(0.88)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e(1.71)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u003cem\u003eGender\u003c/em\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e0.058\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e-0.082\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e0.074*\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e(1.46)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e(-1.30)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e(1.92)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u003cem\u003eGrade\u003c/em\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e0.019\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e0.010\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e0.017\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e(1.10)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e(0.37)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e(1.03)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u003cem\u003esubjBkg\u003c/em\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e0.005\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e0.018\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e0.001\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e(0.12)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e(0.28)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e(0.04)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u003cem\u003einfoAssess\u003c/em\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e0.189***\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e(4.40)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e_cons\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e1.844***\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e1.349***\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e1.588***\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e(10.35)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e(4.81)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e(8.82)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eSobel test\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colspan=\"3\" nameend=\"c4\" namest=\"c2\"\u003e \u003cp\u003e0.102*** (z\u0026thinsp;=\u0026thinsp;3.790)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eGoodman-1 test\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colspan=\"3\" nameend=\"c4\" namest=\"c2\"\u003e \u003cp\u003e0.102*** (z\u0026thinsp;=\u0026thinsp;3.811)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eGoodman-2 test\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colspan=\"3\" nameend=\"c4\" namest=\"c2\"\u003e \u003cp\u003e0.102*** (z\u0026thinsp;=\u0026thinsp;3.770)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eIndirect effect coefficient\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colspan=\"3\" nameend=\"c4\" namest=\"c2\"\u003e \u003cp\u003e0.102*** (z\u0026thinsp;=\u0026thinsp;3.790)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eDirect effect coefficient\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colspan=\"3\" nameend=\"c4\" namest=\"c2\"\u003e \u003cp\u003e0.401*** (z\u0026thinsp;=\u0026thinsp;8.709)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eTotal effect coefficient\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colspan=\"3\" nameend=\"c4\" namest=\"c2\"\u003e \u003cp\u003e0.503*** (z\u0026thinsp;=\u0026thinsp;12.336)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eProportion of intermediary effects\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colspan=\"3\" nameend=\"c4\" namest=\"c2\"\u003e \u003cp\u003e0.203\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eBootstrap (ind_eff)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colspan=\"3\" nameend=\"c4\" namest=\"c2\"\u003e \u003cp\u003e0.102***\u003c/p\u003e \u003cp\u003e[0.052,0.152]\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eBootstrap (dir_eff)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colspan=\"3\" nameend=\"c4\" namest=\"c2\"\u003e \u003cp\u003e0.401***\u003c/p\u003e \u003cp\u003e[0.286,0.516]\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eN\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e205\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e205\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e205\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eR\u0026sup2;\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e0.491\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e0.324\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e0.537\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003c/tbody\u003e \u003c/colgroup\u003e \u003c/table\u003e\u003c/div\u003e \u003c/p\u003e \u003cp\u003eTable\u0026nbsp;\u003cspan refid=\"Tab7\" class=\"InternalRef\"\u003e7\u003c/span\u003e presents the regression results considering the mediating variable \u003cem\u003einfoAssess\u003c/em\u003e. Prior to including the mediating variable \u003cem\u003einfoAssess\u003c/em\u003e, the coefficient of \u003cem\u003eaiSkill\u003c/em\u003e on \u003cem\u003esustEdu\u003c/em\u003e was 0.515 (p\u0026thinsp;\u0026lt;\u0026thinsp;0.01). After introducing the mediating variable, the coefficient decreased to 0.403 (p\u0026thinsp;\u0026lt;\u0026thinsp;0.01), still indicating a significant positive effect, but with reduced effect size. This suggests that \u003cem\u003einfoAssess\u003c/em\u003e partially explains the relationship between \u003cem\u003eaiSkill\u003c/em\u003e and \u003cem\u003esustEdu\u003c/em\u003e. The Sobel test and Goodman test confirmed the significance of the mediation effect of \u003cem\u003einfoAssess\u003c/em\u003e. Additionally, Bootstrap analysis further validated this finding, with a 95% confidence interval [0.286, 0.516] that does not include 0, indicating that the mediating variable's role in transmitting the effect from the independent variable to the dependent variable is stable and not due to random error or spurious association. Thus, \u003cem\u003eaiSkill\u003c/em\u003e indirectly promotes \u003cem\u003esustEdu\u003c/em\u003e by enhancing \u003cem\u003einfoAssess\u003c/em\u003e.\u003c/p\u003e \u003cp\u003eOverall, the mediation effect of \u003cem\u003einfoAssess\u003c/em\u003e accounts for approximately 20% of the total effect (0.203), indicating that information assessment capability explains about 20% of the relationship between the application of Large-Scale Artificial Intelligence Models in library information retrieval systems and ESD. This suggests that merely implementing Large-Scale Artificial Intelligence Models in library information retrieval systems is insufficient to achieve education goals for sustainable development; it is also necessary to concurrently strengthen individuals' information assessment capabilities.\u003c/p\u003e \u003cp\u003e \u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"Yes\" id=\"Tab8\" border=\"1\"\u003e \u003ccaption language=\"En\"\u003e \u003cdiv class=\"CaptionNumber\"\u003eTable 8\u003c/div\u003e \u003cdiv class=\"CaptionContent\"\u003e \u003cp\u003eModel regression result (\u003cem\u003edigLit\u003c/em\u003e).\u003c/p\u003e \u003c/div\u003e \u003c/caption\u003e \u003ccolgroup cols=\"4\"\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c3\" colnum=\"3\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c4\" colnum=\"4\"\u003e\u003c/div\u003e \u003cthead\u003e \u003ctr\u003e \u003cth align=\"left\" colname=\"c1\" morerows=\"1\" rowspan=\"2\"\u003e\u0026nbsp;\u003c/th\u003e \u003cth align=\"left\" colname=\"c2\"\u003e \u003cp\u003e(1)\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c3\"\u003e \u003cp\u003e(2)\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c4\"\u003e \u003cp\u003e(3)\u003c/p\u003e \u003c/th\u003e \u003c/tr\u003e \u003ctr\u003e \u003cth align=\"left\" colname=\"c2\"\u003e \u003cp\u003e\u003cem\u003esustEdu\u003c/em\u003e\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c3\"\u003e \u003cp\u003e\u003cem\u003edigLit\u003c/em\u003e\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c4\"\u003e \u003cp\u003e\u003cem\u003esustEdu\u003c/em\u003e\u003c/p\u003e \u003c/th\u003e \u003c/tr\u003e \u003c/thead\u003e \u003ctbody\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u003cem\u003eaiSkill\u003c/em\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e0.515***\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e0.815***\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e0.136**\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e(12.98)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e(20.37)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e(2.21)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u003cem\u003eprov\u003c/em\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e0.076*\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e0.036\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e0.060*\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e(1.90)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e(0.89)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e(1.67)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u003cem\u003eGender\u003c/em\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e0.058\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e0.068*\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e0.027\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e(1.46)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e(1.68)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e(0.75)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u003cem\u003eGrade\u003c/em\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e0.019\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e-0.004\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e0.021\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e(1.10)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e(-0.25)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e(1.37)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u003cem\u003esubjBkg\u003c/em\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e0.005\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e0.043\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e-0.015\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e(0.12)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e(1.07)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e(-0.43)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u003cem\u003edigLit\u003c/em\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e0.465***\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e(7.47)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e_cons\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e1.844***\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e0.567***\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e1.580***\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e(10.35)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e(3.16)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e(9.78)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eSobel test\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colspan=\"3\" nameend=\"c4\" namest=\"c2\"\u003e \u003cp\u003e0.366*** (z\u0026thinsp;=\u0026thinsp;6.842)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eGoodman-1 test\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colspan=\"3\" nameend=\"c4\" namest=\"c2\"\u003e \u003cp\u003e0.366*** (z\u0026thinsp;=\u0026thinsp;6.850)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eGoodman-2 test\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colspan=\"3\" nameend=\"c4\" namest=\"c2\"\u003e \u003cp\u003e0.366*** (z\u0026thinsp;=\u0026thinsp;6.834)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eIndirect effect coefficient\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colspan=\"3\" nameend=\"c4\" namest=\"c2\"\u003e \u003cp\u003e0.366*** (z\u0026thinsp;=\u0026thinsp;6.842)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eDirect effect coefficient\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colspan=\"3\" nameend=\"c4\" namest=\"c2\"\u003e \u003cp\u003e0.136*** (z\u0026thinsp;=\u0026thinsp;2.201)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eTotal effect coefficient\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colspan=\"3\" nameend=\"c4\" namest=\"c2\"\u003e \u003cp\u003e0.503*** (z\u0026thinsp;=\u0026thinsp;12.336)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eProportion of intermediary effects\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colspan=\"3\" nameend=\"c4\" namest=\"c2\"\u003e \u003cp\u003e0.729\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eBootstrap (ind_eff)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colspan=\"3\" nameend=\"c4\" namest=\"c2\"\u003e \u003cp\u003e0.366***\u003c/p\u003e \u003cp\u003e[0.255,0.478]\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eBootstrap (dir_eff)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colspan=\"3\" nameend=\"c4\" namest=\"c2\"\u003e \u003cp\u003e0.136*\u003c/p\u003e \u003cp\u003e[-0.009,0.281]\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eN\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e205\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e205\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e205\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eR\u0026sup2;\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e0.491\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e0.690\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e0.603\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003c/tbody\u003e \u003c/colgroup\u003e \u003c/table\u003e\u003c/div\u003e \u003c/p\u003e \u003cp\u003eTable\u0026nbsp;\u003cspan refid=\"Tab8\" class=\"InternalRef\"\u003e8\u003c/span\u003e presents the regression results considering the mediating variable \u003cem\u003edigLit\u003c/em\u003e. Similarly, after introducing \u003cem\u003edigLit\u003c/em\u003e, the effect of \u003cem\u003eaiSkill\u003c/em\u003e on \u003cem\u003esustEdu\u003c/em\u003e changed, with the coefficient decreasing from 0.515 to 0.136, and the significance level shifting from significant at the 1% level to significant at the 5% level. This preliminary finding suggests that \u003cem\u003edigLit\u003c/em\u003e mediates the relationship between \u003cem\u003eaiSkill\u003c/em\u003e and \u003cem\u003esustEdu\u003c/em\u003e. Subsequent Sobel and Goodman tests confirmed the significance of the mediation effect of \u003cem\u003edigLit\u003c/em\u003e (z\u0026thinsp;=\u0026thinsp;6.842\u0026ndash;6.850, p\u0026thinsp;\u0026lt;\u0026thinsp;0.01), with a mediation effect coefficient of 0.366. Bootstrap analysis further validated this result, with a 95% confidence interval [0.255, 0.478] that does not include 0, confirming that \u003cem\u003eaiSkill\u003c/em\u003e indirectly influences \u003cem\u003esustEdu\u003c/em\u003e through \u003cem\u003edigLit\u003c/em\u003e.\u003c/p\u003e \u003cp\u003eOverall, the total effect coefficient is 0.503 (z\u0026thinsp;=\u0026thinsp;12.336, p\u0026thinsp;\u0026lt;\u0026thinsp;0.01), with the mediation effect of \u003cem\u003edigLit\u003c/em\u003e accounting for approximately 73% (0.729) of the total effect, which is substantially higher than the mediation effect of \u003cem\u003einfoAssess\u003c/em\u003e.\u003c/p\u003e \u003cp\u003e \u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"Yes\" id=\"Tab9\" border=\"1\"\u003e \u003ccaption language=\"En\"\u003e \u003cdiv class=\"CaptionNumber\"\u003eTable 9\u003c/div\u003e \u003cdiv class=\"CaptionContent\"\u003e \u003cp\u003eModel regression result (\u003cem\u003epvtSec\u003c/em\u003e).\u003c/p\u003e \u003c/div\u003e \u003c/caption\u003e \u003ccolgroup cols=\"4\"\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c3\" colnum=\"3\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c4\" colnum=\"4\"\u003e\u003c/div\u003e \u003cthead\u003e \u003ctr\u003e \u003cth align=\"left\" colname=\"c1\" morerows=\"1\" rowspan=\"2\"\u003e\u0026nbsp;\u003c/th\u003e \u003cth align=\"left\" colname=\"c2\"\u003e \u003cp\u003e(1)\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c3\"\u003e \u003cp\u003e(2)\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c4\"\u003e \u003cp\u003e(3)\u003c/p\u003e \u003c/th\u003e \u003c/tr\u003e \u003ctr\u003e \u003cth align=\"left\" colname=\"c2\"\u003e \u003cp\u003e\u003cem\u003esustEdu\u003c/em\u003e\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c3\"\u003e \u003cp\u003e\u003cem\u003epvtSec\u003c/em\u003e\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c4\"\u003e \u003cp\u003e\u003cem\u003esustEdu\u003c/em\u003e\u003c/p\u003e \u003c/th\u003e \u003c/tr\u003e \u003c/thead\u003e \u003ctbody\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u003cem\u003eaiSkill\u003c/em\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e0.515***\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e0.347***\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e0.463***\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e(12.98)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e(4.41)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e(11.64)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u003cem\u003eprov\u003c/em\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e0.076*\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e-0.015\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e0.079**\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e(1.90)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e(-0.18)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e(2.05)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u003cem\u003eGender\u003c/em\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e0.058\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e0.105\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e0.042\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e(1.46)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e(1.33)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e(1.11)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u003cem\u003eGrade\u003c/em\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e0.019\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e-0.052\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e0.027\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e(1.10)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e(-1.55)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e(1.62)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u003cem\u003esubjBkg\u003c/em\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e0.005\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e0.045\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e-0.002\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e(0.12)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e(0.56)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e(-0.05)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u003cem\u003epvtSec\u003c/em\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e0.150***\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e(4.40)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e_cons\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e1.844***\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e3.215***\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e1.360***\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e(10.35)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e(9.10)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e(6.71)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eSobel test\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colspan=\"3\" nameend=\"c4\" namest=\"c2\"\u003e \u003cp\u003e0.055*** (z\u0026thinsp;=\u0026thinsp;3.148)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eGoodman-1 test\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colspan=\"3\" nameend=\"c4\" namest=\"c2\"\u003e \u003cp\u003e0.055*** (z\u0026thinsp;=\u0026thinsp;3.188)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eGoodman-2 test\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colspan=\"3\" nameend=\"c4\" namest=\"c2\"\u003e \u003cp\u003e0.055*** (z\u0026thinsp;=\u0026thinsp;3.109)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eIndirect effect coefficient\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colspan=\"3\" nameend=\"c4\" namest=\"c2\"\u003e \u003cp\u003e0.055*** (z\u0026thinsp;=\u0026thinsp;3.148)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eDirect effect coefficient\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colspan=\"3\" nameend=\"c4\" namest=\"c2\"\u003e \u003cp\u003e0.448*** (z\u0026thinsp;=\u0026thinsp;10.979)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eTotal effect coefficient\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colspan=\"3\" nameend=\"c4\" namest=\"c2\"\u003e \u003cp\u003e0.503*** (z\u0026thinsp;=\u0026thinsp;12.336)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eProportion of intermediary effects\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colspan=\"3\" nameend=\"c4\" namest=\"c2\"\u003e \u003cp\u003e0.109\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eBootstrap (ind_eff)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colspan=\"3\" nameend=\"c4\" namest=\"c2\"\u003e \u003cp\u003e0.055***\u003c/p\u003e \u003cp\u003e[0.020,0.089]\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eBootstrap (dir_eff)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colspan=\"3\" nameend=\"c4\" namest=\"c2\"\u003e \u003cp\u003e0.448***\u003c/p\u003e \u003cp\u003e[0.346,0.550]\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eN\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e205\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e205\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e205\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eR\u0026sup2;\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e0.491\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e0.104\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e0.537\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003c/tbody\u003e \u003c/colgroup\u003e \u003c/table\u003e\u003c/div\u003e \u003c/p\u003e \u003cp\u003eTable\u0026nbsp;\u003cspan refid=\"Tab9\" class=\"InternalRef\"\u003e9\u003c/span\u003e presents the regression results considering the mediating variable \u003cem\u003epvtSec\u003c/em\u003e. Similarly, after introducing \u003cem\u003epvtSec\u003c/em\u003e, the coefficient of \u003cem\u003eaiSkill\u003c/em\u003e on \u003cem\u003esustEdu\u003c/em\u003e changed from 0.515 to 0.463, with no change in statistical significance. The Sobel and Goodman tests confirmed the significance of the mediation effect of \u003cem\u003epvtSec\u003c/em\u003e (z\u0026thinsp;=\u0026thinsp;3.148\u0026ndash;3.109, p\u0026thinsp;\u0026lt;\u0026thinsp;0.01), with a mediation effect coefficient of 0.055. Bootstrap analysis further validated this finding, with a 95% confidence interval [0.020, 0.089] that does not include 0, confirming that \u003cem\u003eaiSkill\u003c/em\u003e indirectly influences \u003cem\u003esustEdu\u003c/em\u003e through \u003cem\u003epvtSec\u003c/em\u003e.\u003c/p\u003e \u003cp\u003eSimilarly, calculations indicate that the mediation effect of \u003cem\u003epvtSec\u003c/em\u003e accounts for approximately 11% (0.109) of the total effect, suggesting a relatively weak but statistically significant mediating role.\u003c/p\u003e \u003cp\u003eThe examination of the mediation effects further validates the hypotheses proposed in the study (1a\u0026ndash;3b). All mediation effects were significant, indicating that the three variables (\u003cem\u003einfoAssess\u003c/em\u003e, \u003cem\u003edigLit\u003c/em\u003e, and \u003cem\u003epvtSec\u003c/em\u003e) each play a mediating role in the pathway through which \u003cem\u003eaiSkill\u003c/em\u003e influences \u003cem\u003esustEdu\u003c/em\u003e. Among these, \u003cem\u003edigLit\u003c/em\u003e exhibited the strongest mediation effect, accounting for approximately 73% of the total effect, while \u003cem\u003epvtSec\u003c/em\u003e had the weakest mediation effect, contributing about 11% to the total effect. Further analysis reveals that \u003cem\u003einfoAssess\u003c/em\u003e supports the knowledge acquisition and application for sustainable education by enhancing individuals' critical evaluation capabilities of information. \u003cem\u003eDigLit\u003c/em\u003e facilitates learners' effective participation in sustainable education practices within digital environments by improving proficiency in operating digital tools. Although the mediation effect of \u003cem\u003epvtSec\u003c/em\u003e is relatively smaller, it still contributes to the ethical foundation of sustainable education by increasing awareness of data privacy and security. These findings not only confirm the driving role of technology use in capability development as posited by the Technology Acceptance Model but also align with the integrated perspective of the artificial intelligence literacy framework and sustainable development theory. They demonstrate that the mediating variables establish critical transmission pathways between technology and educational outcomes.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec20\" class=\"Section2\"\u003e \u003ch2\u003e4.6. Robustness test\u003c/h2\u003e \u003cp\u003e \u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"Yes\" id=\"Tab10\" border=\"1\"\u003e \u003ccaption language=\"En\"\u003e \u003cdiv class=\"CaptionNumber\"\u003eTable 10\u003c/div\u003e \u003cdiv class=\"CaptionContent\"\u003e \u003cp\u003eRobustness test.\u003c/p\u003e \u003c/div\u003e \u003c/caption\u003e \u003ccolgroup cols=\"3\"\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c3\" colnum=\"3\"\u003e\u003c/div\u003e \u003cthead\u003e \u003ctr\u003e \u003cth align=\"left\" colname=\"c1\" morerows=\"1\" rowspan=\"2\"\u003e\u0026nbsp;\u003c/th\u003e \u003cth align=\"left\" colname=\"c2\"\u003e \u003cp\u003e(1)\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c3\"\u003e \u003cp\u003e(2)\u003c/p\u003e \u003c/th\u003e \u003c/tr\u003e \u003ctr\u003e \u003cth align=\"left\" colname=\"c2\"\u003e \u003cp\u003e\u003cem\u003esustEdu\u003c/em\u003e\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c3\"\u003e \u003cp\u003e\u003cem\u003esustEdu\u003c/em\u003e\u003c/p\u003e \u003c/th\u003e \u003c/tr\u003e \u003c/thead\u003e \u003ctbody\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u003cem\u003eaiSkill\u003c/em\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e0.503***\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e(12.34)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u003cem\u003eaiFreq\u003c/em\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e0.088***\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e(2.98)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u003cem\u003eprov\u003c/em\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e0.076*\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e0.081\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e(1.90)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e(1.55)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u003cem\u003eGender\u003c/em\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e0.057\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e0.106**\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e(1.44)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e(2.05)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u003cem\u003eGrade\u003c/em\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e0.018\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e0.035\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e(1.03)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e(1.53)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u003cem\u003esubjBkg\u003c/em\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e-0.001\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e0.006\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e(-0.03)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e(0.11)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u003cem\u003eaiEduLevel\u003c/em\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e0.041\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e0.105**\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e(1.30)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e(2.56)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e_cons\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e1.817***\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e3.352***\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e(10.15)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e(21.13)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eN\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e205\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e205\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eR\u0026sup2;\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e0.496\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e0.146\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colspan=\"3\" nameend=\"c3\" namest=\"c1\"\u003e \u003cp\u003eNote: t statistics in parentheses,\u003c/p\u003e \u003cp\u003e* p\u0026thinsp;\u0026lt;\u0026thinsp;0.1, ** p\u0026thinsp;\u0026lt;\u0026thinsp;0.05, *** p\u0026thinsp;\u0026lt;\u0026thinsp;0.01\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003c/tbody\u003e \u003c/colgroup\u003e \u003c/table\u003e\u003c/div\u003e \u003c/p\u003e \u003cp\u003eThe study conducted a robustness test by replacing the explanatory variable, substituting the proficiency in using library information retrieval systems (\u003cem\u003eaiSkill\u003c/em\u003e) with the frequency of using library AI information retrieval systems (\u003cem\u003eaiFreq\u003c/em\u003e). The results are presented in Table\u0026nbsp;\u003cspan refid=\"Tab10\" class=\"InternalRef\"\u003e10\u003c/span\u003e. Specifically, the original variable \u003cem\u003eaiSkill\u003c/em\u003e (representing the use of Large-Scale Artificial Intelligence Models in library information retrieval systems) was replaced with \u003cem\u003eaiFreq\u003c/em\u003e (representing the frequency of using library AI information retrieval systems), and regression analysis was performed to assess the robustness of the model.\u003c/p\u003e \u003cp\u003eThe results indicate that the direct effect coefficient of \u003cem\u003eaiSkill\u003c/em\u003e on \u003cem\u003esustEdu\u003c/em\u003e in Model (1) was 0.503 (p\u0026thinsp;\u0026lt;\u0026thinsp;0.01). In Model (2), after replacing the variable, the effect of \u003cem\u003eaiFreq\u003c/em\u003e on \u003cem\u003esustEdu\u003c/em\u003e remained significant, with a coefficient of 0.088 (p\u0026thinsp;\u0026lt;\u0026thinsp;0.01), demonstrating that the frequency of use has a significant positive impact on ESD.\u003c/p\u003e \u003cp\u003eThe robustness test results suggest that, although the substitution of variables led to changes in the magnitude of the effect, the positive impact of \u003cem\u003eaiSkill\u003c/em\u003e on \u003cem\u003esustEdu\u003c/em\u003e remained consistent across different measurement dimensions, confirming the reliability of the model results. This finding supports the stability of the research hypotheses and indicates that usage frequency and usage proficiency may contribute differently to explaining the effects on sustainable education, warranting further exploration in future research.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec21\" class=\"Section2\"\u003e \u003ch2\u003e4.7. Heterogeneity test\u003c/h2\u003e \u003cp\u003e \u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"Yes\" id=\"Tab11\" border=\"1\"\u003e \u003ccaption language=\"En\"\u003e \u003cdiv class=\"CaptionNumber\"\u003eTable 11\u003c/div\u003e \u003cdiv class=\"CaptionContent\"\u003e \u003cp\u003eHeterogeneity test.\u003c/p\u003e \u003c/div\u003e \u003c/caption\u003e \u003ccolgroup cols=\"7\"\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c3\" colnum=\"3\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c4\" colnum=\"4\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c5\" colnum=\"5\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c6\" colnum=\"6\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c7\" colnum=\"7\"\u003e\u003c/div\u003e \u003cthead\u003e \u003ctr\u003e \u003cth align=\"left\" colname=\"c1\" morerows=\"1\" rowspan=\"2\"\u003e\u0026nbsp;\u003c/th\u003e \u003cth align=\"left\" colname=\"c2\"\u003e \u003cp\u003e(1) developed regions\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c3\"\u003e \u003cp\u003e(2) developing regions\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c4\"\u003e \u003cp\u003e(3) Female\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c5\"\u003e \u003cp\u003e(4) Male\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c6\"\u003e \u003cp\u003e(5) Low AI Education Level\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c7\"\u003e \u003cp\u003e(6) High AI Education Level\u003c/p\u003e \u003c/th\u003e \u003c/tr\u003e \u003ctr\u003e \u003cth align=\"left\" colname=\"c2\"\u003e \u003cp\u003e\u003cem\u003esustEdu\u003c/em\u003e\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c3\"\u003e \u003cp\u003e\u003cem\u003esustEdu\u003c/em\u003e\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c4\"\u003e \u003cp\u003e\u003cem\u003esustEdu\u003c/em\u003e\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c5\"\u003e \u003cp\u003e\u003cem\u003esustEdu\u003c/em\u003e\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c6\"\u003e \u003cp\u003e\u003cem\u003esustEdu\u003c/em\u003e\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c7\"\u003e \u003cp\u003e\u003cem\u003esustEdu\u003c/em\u003e\u003c/p\u003e \u003c/th\u003e \u003c/tr\u003e \u003c/thead\u003e \u003ctbody\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u003cem\u003eaiSkill\u003c/em\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e0.605***\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e0.449***\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e0.386***\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e0.594***\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e0.471***\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e0.556***\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e(8.51)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e(8.82)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e(6.60)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e(10.42)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e(9.60)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e(8.59)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u003cem\u003eGender\u003c/em\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e0.067\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e0.054\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e0.067\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e0.065\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e(1.13)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e(0.99)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e(1.33)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e(1.20)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u003cem\u003eGrade\u003c/em\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e-0.002\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e0.033\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e0.018\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e0.020\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e0.032\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e-0.021\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e(-0.07)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e(1.42)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e(0.77)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e(0.82)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e(1.50)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e(-0.83)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u003cem\u003esubjBkg\u003c/em\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e0.003\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e0.012\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e0.026\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e-0.036\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e0.013\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e-0.041\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e(0.05)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e(0.23)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e(0.49)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e(-0.59)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e(0.26)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e(-0.71)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u003cem\u003eaiEduLevel\u003c/em\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e0.067\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e0.009\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e0.082\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e0.012\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e(1.30)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e(0.21)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e(1.63)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e(0.30)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u003cem\u003eprov\u003c/em\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e0.067\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e0.103\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e0.101**\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e0.009\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e(1.29)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e(1.64)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e(2.04)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e(0.16)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e_cons\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e1.467***\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e2.198***\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e2.236***\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e1.519***\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e1.897***\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e2.033***\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e(5.20)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e(10.03)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e(9.01)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e(5.76)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e(8.65)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e(6.74)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eN\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e84\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e121\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e111\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e94\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e148\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e57\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eR\u0026sup2;\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e0.577\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e0.453\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e0.377\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e0.587\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e0.449\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e0.606\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003c/tbody\u003e \u003c/colgroup\u003e \u003c/table\u003e\u003c/div\u003e \u003c/p\u003e \u003cp\u003eThe study further conducted a heterogeneity analysis to examine the varying effects of \u003cem\u003eaiSkill\u003c/em\u003e on \u003cem\u003esustEdu\u003c/em\u003e across regions with different levels of economic development, groups of different genders, and groups with varying levels of artificial intelligence education. The results are presented in Table\u0026nbsp;\u003cspan refid=\"Tab11\" class=\"InternalRef\"\u003e11\u003c/span\u003e.\u003c/p\u003e \u003cp\u003eHeterogeneity Analysis Across Regions with Different Levels of Economic Development. To further explore the heterogeneity in the effect of \u003cem\u003eaiSkill\u003c/em\u003e on \u003cem\u003esustEdu\u003c/em\u003e, the study first examined the differential effects across regions with varying levels of economic development. In regions with higher economic development (Column 1), the regression coefficient of \u003cem\u003eaiSkill\u003c/em\u003e on \u003cem\u003esustEdu\u003c/em\u003e was 0.605 (p\u0026thinsp;\u0026lt;\u0026thinsp;0.01), with an R\u0026sup2; of 0.577. In contrast, in regions with lower economic development (Column 2), the coefficient decreased to 0.449 (p\u0026thinsp;\u0026lt;\u0026thinsp;0.01), with an R\u0026sup2; of 0.453. These results indicate that the promoting effect of \u003cem\u003eaiSkill\u003c/em\u003e on sustainable development education is more pronounced in economically developed regions. In economically developed regions, well-established hardware infrastructure ensures the stable operation of educational applications related to \u003cem\u003eaiSkill\u003c/em\u003e. Abundant data resources support the optimization of algorithms and the customization of personalized teaching plans. An open and innovative educational ecosystem, coupled with teachers' higher digital literacy, facilitates the rapid integration of \u003cem\u003eaiSkill\u003c/em\u003e into sustainable development education classrooms. Additionally, strong societal demand for talent and ample policy and financial support further promote the deep integration of \u003cem\u003eaiSkill\u003c/em\u003e and sustainable education. The convergence of these favorable conditions amplifies the positive impact of \u003cem\u003eaiSkill\u003c/em\u003e on ESD. This heterogeneity reflects the adaptive differences in the application of Large-Scale Artificial Intelligence Models in library information retrieval systems across diverse socioeconomic contexts. It further suggests that future educational interventions should be tailored to regional characteristics to address these disparities effectively.\u003c/p\u003e \u003cp\u003eHeterogeneity Analysis Across Gender Groups. The study also analyzed the heterogeneity in the effect of \u003cem\u003eaiSkill\u003c/em\u003e on \u003cem\u003esustEdu\u003c/em\u003e across different gender groups. The results indicate that in the female group (Column 3), the regression coefficient of \u003cem\u003eaiSkill\u003c/em\u003e on \u003cem\u003esustEdu\u003c/em\u003e was 0.386 (p\u0026thinsp;\u0026lt;\u0026thinsp;0.01), with an R\u0026sup2; of 0.377. In the male group (Column 4), the coefficient was 0.594 (p\u0026thinsp;\u0026lt;\u0026thinsp;0.01), with an R\u0026sup2; of 0.587. Although both effects are significant, the effect is slightly weaker in the female group, which may be related to differences in learning strategies and personality tendencies. Gender differences are influenced by sociocultural norms or educational support systems. To narrow this impact gap, the participation of females in artificial intelligence skills training should be further enhanced through targeted interventions.\u003c/p\u003e \u003cp\u003eHeterogeneity Analysis Across Groups with Different Levels of AI Education. Additionally, the study examined the heterogeneity in the effect of \u003cem\u003eaiSkill\u003c/em\u003e on \u003cem\u003esustEdu\u003c/em\u003e across groups with varying levels of artificial intelligence education. The results indicate that in the group with a low level of AI education (Column 5), the coefficient of \u003cem\u003eaiSkill\u003c/em\u003e on \u003cem\u003esustEdu\u003c/em\u003e was 0.471 (p\u0026thinsp;\u0026lt;\u0026thinsp;0.01), with an R\u0026sup2; of 0.449. In contrast, in the group with a high level of AI education (Column 6), the coefficient increased to 0.556 (p\u0026thinsp;\u0026lt;\u0026thinsp;0.01), with an R\u0026sup2; of 0.606. The effect size in the high-level AI education group was significantly stronger than in the low-level group, reflecting the advantages of advanced education in deepening skills and broadening application scope, which enhances its promoting effect on sustainable development education. The increase in R\u0026sup2; values indicates that the model has greater explanatory power in the high-level AI education group.\u003c/p\u003e \u003cp\u003eThis heterogeneity stems from the cumulative effect of education level, where high-level AI education not only enhances technical proficiency but also strengthens learners' ability to translate skills into educational outcomes in complex contexts. However, the effect in the low-level group, though weaker, remains significant, suggesting that basic AI skills still hold practical value at the initial stage. These findings recommend a tiered approach to AI education: for low-level groups, foundational application training should be strengthened, while high-level groups should receive support for advanced projects to maximize the overall impact of \u003cem\u003eaiSkill\u003c/em\u003e on \u003cem\u003esustEdu\u003c/em\u003e.\u003c/p\u003e \u003c/div\u003e"},{"header":"5. Conclusions","content":"\u003cp\u003eThis study investigates the impact of applying Large-Scale Artificial Intelligence Models to library information retrieval systems on ESD, the pathways of this impact, and its heterogeneity across different gender groups, regions, and groups with varying levels of AI education.\u003c/p\u003e \u003cp\u003eThe Technology Acceptance Model, the artificial intelligence literacy framework, and sustainable development theory collectively form an integrated model to explain the effects on ESD. Compared to individual theories, the results demonstrate that the integrated model offers greater explanatory power. From the perspective of the Technology Acceptance Model, users' perceived usefulness and ease of use of library AI retrieval systems drive their usage behavior, which in turn influences their engagement with sustainable development education. Within the artificial intelligence literacy framework, information assessment capability, digital literacy, and privacy and security awareness serve as key mediators: information assessment capability enables users to accurately filter information relevant to ESD, digital literacy ensures efficient use of the system to access resources, and privacy and security awareness establishes a foundation of trust in system usage. Sustainable development theory, from a macro perspective, explains how education promotes the coordinated development of society, the environment, and the economy, providing direction for the impact pathways. Compared to a single theory, this integrated approach mitigates the Technology Acceptance Model's limited focus on users' intrinsic competencies, compensates for the artificial intelligence literacy framework's lack of macro-level value articulation, and addresses the shortcomings of sustainable development theory in elucidating micro-level pathways for technology application.\u003c/p\u003e \u003cdiv id=\"Sec23\" class=\"Section2\"\u003e \u003ch2\u003e5.1.Recommendations\u003c/h2\u003e \u003cp\u003eBased on the findings that the application of Large-Scale Artificial Intelligence Models in library information retrieval systems can facilitate the achievement of educational goals for sustainable development, the study proposes the following recommendations:\u003c/p\u003e \u003cp\u003ePolicy Recommendations: Government and educational authorities should promote the localized deployment of Large-Scale Artificial Intelligence Models in higher education institutions and advance the digitization of university and local libraries through policy support and funding to enhance the realization of ESD objectives. The study demonstrates that the application of library AI retrieval systems significantly influences ESD through information assessment capability, digital literacy, and privacy and security awareness, with \u003cem\u003edigLit\u003c/em\u003e accounting for 73% of the total mediation effect (Table\u0026nbsp;\u003cspan refid=\"Tab8\" class=\"InternalRef\"\u003e8\u003c/span\u003e). To this end, policies could draw on the European Union's Digital Education Action Plan (2021\u0026ndash;2027) by allocating dedicated funds and implementing training programs to support the integration of AI technologies in education [\u003cspan citationid=\"CR35\" class=\"CitationRef\"\u003e35\u003c/span\u003e]. Educational policies and frameworks should emphasize the development of students' digital literacy and information assessment capabilities, which are critical for the interdisciplinary learning required in ESD [\u003cspan citationid=\"CR36\" class=\"CitationRef\"\u003e36\u003c/span\u003e]. Furthermore, the heterogeneity analysis reveals that the effect of \u003cem\u003eaiSkill\u003c/em\u003e is stronger in economically developed regions, among males, and in groups with higher levels of AI education, suggesting that policies should optimize resource allocation by accounting for regional and individual differences. Additionally, stringent data privacy and security standards should be established to ensure compliance and legality in the application of Large-Scale Artificial Intelligence Models in educational settings, thereby enhancing users' trust in AI systems and improving usage willingness and educational outcomes.\u003c/p\u003e \u003cp\u003eEducational Practice. Based on the research findings, higher education institutions should integrate knowledge about AI information retrieval systems, applications of Large-Scale Artificial Intelligence Models, and related literacy into curriculum design to foster students' sustainable development awareness and technical skills. The study demonstrates that the use of library AI information retrieval systems promotes \u003cem\u003esustEdu\u003c/em\u003e by enhancing \u003cem\u003edigLit\u003c/em\u003e, \u003cem\u003einfoAssess\u003c/em\u003e, and \u003cem\u003epvtSec\u003c/em\u003e. Educators should receive professional training to effectively guide students in using AI tools within library settings. Additionally, the evaluation systems for higher education should be updated to include students' abilities to address sustainable development issues using AI tools, thereby comprehensively assessing their learning outcomes. Furthermore, the robustness test indicates that the effect remains significant even after variable substitution, underscoring that educational practices should prioritize skill development rather than merely relying on usage frequency. This approach not only supports the enhancement of perceived usefulness and ease of use as outlined in the Technology Acceptance Model, but also aligns with the artificial intelligence literacy framework's goals of 'applying AI' and 'evaluating and creating AI'.\u003c/p\u003e \u003cp\u003eInternational Experience. A recent report from the U.S. Department of Education, 'Artificial Intelligence and the Future of Teaching and Learning', provides valuable insights into the role of artificial intelligence in education and its implications for various stakeholders, including academic libraries. International experiences offer valuable references for the application of Large-Scale Artificial Intelligence Models in Chinese university libraries. In the United Kingdom, university libraries have enhanced research efficiency through AI tools, such as the AI recommendation system implemented by the University of London Library. This successful experience can serve as a model for Chinese higher education institutions. The European Union promotes AI literacy education through cross-national collaborative projects, emphasizing the cultivation of ethical awareness, which aligns with the findings in this study that Chinese universities need to strengthen \u003cem\u003epvtSec\u003c/em\u003e. In contrast, developing countries like India have expanded access to educational resources through low-cost AI solutions, such as voice assistants, offering a reference for less developed regions in China. By integrating these experiences, China should formulate localized strategies that balance technological advancement with educational equity to achieve sustainable development goals.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec24\" class=\"Section2\"\u003e \u003ch2\u003e5.2. Limitations and future work\u003c/h2\u003e \u003cp\u003eAlthough this study systematically analyzed the promoting effect of library AI information retrieval system usage on ESD and its mediating mechanisms through the Technology Acceptance Model, the artificial intelligence literacy framework, and sustainable development theory, several limitations remain. First, the sample was primarily composed of students from Chinese higher education institutions, and the homogeneity in geographic and cultural backgrounds limits the generalizability of the findings. Future research could expand to include international multicenter samples to compare the impact of Large-Scale Artificial Intelligence Models usage in library information retrieval systems on ESD across diverse cultural contexts. Second, the study relied on self-reported data and cross-sectional analysis, which could not capture the dynamic changes in mediating variables resulting from long-term usage behavior. Future studies could employ longitudinal research or experimental designs to validate causal relationships. Third, the study did not thoroughly explore the technical characteristics or ethical issues of AI systems. Future research could incorporate technical evaluation frameworks to analyze their potential negative effects.\u003c/p\u003e \u003c/div\u003e"},{"header":"Declarations","content":"\u003ch2\u003eConflicts of Interest:\u003c/h2\u003e\n\u003cp\u003eThe authors declare no competing interests.\u003c/p\u003e\n\u003ch2\u003eFunding:\u003c/h2\u003e\n\u003cp\u003eThis research received no external funding.\u003c/p\u003e\n\u003ch2\u003eAuthor Contribution\u003c/h2\u003e\n\u003cp\u003eConceptualization, Y.Z., Z.L., and C.W.; data curation, Y.Z. and Z.L.; formal analysis, Y.Z., Z.L., and C.W.; investigation, Y.Z., C.W., and Z.L.; methodology, Y.Z., C.W., and Z.L.; project administration, C.W.; resources, Y.Z., Z.L., and C.W.; supervision, C.W.; validation, Y.Z., Z.L., and C.W.; visualization, Z.L.; writing\u0026mdash;original draft, Y.Z. and Z.L.; writing\u0026mdash;review and editing, Y.Z., Z.L., and C.W.. All the authors have read and agreed to the published version of the manuscript.\u003c/p\u003e\n\u003ch2\u003eData Availability\u003c/h2\u003e\n\u003cp\u003eThe data presented in this study are available on request from the corresponding author due to questionnaire respondents\u0026apos; confidentiality.\u003c/p\u003e"},{"header":"References","content":"\u003col\u003e\u003cli\u003e\u003cspan\u003eUnited Nations General Assembly. Education for sustainable development in the framework of the 2030 Agenda for Sustainable Development: Report of the Director-General of UNESCO (A/78/219). (2023). Available at: \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://docs.un.org/en/A/78/219\u003c/span\u003e\u003cspan address=\"https://docs.un.org/en/A/78/219\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eGiannini, S. Education for sustainable development\u0026mdash;Roadmap. \u003cem\u003eUnited Nations Educational, Scientific and Cultural Organization\u003c/em\u003e (2020). \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.54675/YFRE1448\u003c/span\u003e\u003cspan address=\"10.54675/YFRE1448\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSingun, A. Jr. Unveiling the barriers to digital transformation in higher education institutions: a systematic literature review. \u003cem\u003eDiscov Educ.\u003c/em\u003e \u003cb\u003e4\u003c/b\u003e, 37 (2025).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eAvelar, A. B. A., Mitchell, S. F. \u0026amp; Sandes-Guimar\u0026atilde;es, L. V. Integrating sustainable development goals in management education: Impact on student knowledge, attitudes, and behaviors. \u003cem\u003eInternational J. Manage. Education\u003c/em\u003e. \u003cb\u003e23\u003c/b\u003e, 101116 (2025).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSavec, V. F. \u0026amp; Jedrinović, S. The Role of AI Implementation in Higher Education in Achieving the Sustainable Development Goals: A Case Study from Slovenia. \u003cem\u003eSustainability\u003c/em\u003e 17, (2024).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eGoralski, M. A. \u0026amp; Tan, T. K. Artificial intelligence and sustainable development. \u003cem\u003eInt. J. Manage. Educ.\u003c/em\u003e \u003cb\u003e18\u003c/b\u003e, 100330 (2020).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eGarnier, M., Cussac, M., Delaune, C. \u0026amp; Espy, C. \u0026amp; Papon-Vidal, L. The Pivotal Role of Libraries in Sustainable AI Development. \u003cem\u003ecjils\u003c/em\u003e 47, 137\u0026ndash;149 (2024).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eKhan, R., Gupta, N., Sinhababu, A. \u0026amp; Chakravarty, R. Impact of Conversational and Generative AI Systems on Libraries: A Use Case Large Language Model (LLM). \u003cem\u003eScience \u0026amp; Technol. Libraries\u003c/em\u003e. \u003cb\u003e43\u003c/b\u003e, 319\u0026ndash;333 (2024).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eInternational Federation of Cartographic Associations. The strategy of the International Federation of Cartographic Associations 2024\u0026ndash;2029. (2024). Available at: \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://www.lsc.org.cn/cns/html_pass/contents/1127/1848974105203838976.html\u003c/span\u003e\u003cspan address=\"https://www.lsc.org.cn/cns/html_pass/contents/1127/1848974105203838976.html\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eKhan, R., Gupta, N., Sinhababu, A. \u0026amp; Chakravarty, R. Impact of Conversational and Generative AI Systems on Libraries: A Use Case Large Language Model (LLM). \u003cem\u003eScience \u0026amp; Technol. Libraries\u003c/em\u003e. \u003cb\u003e43\u003c/b\u003e, 319\u0026ndash;333 (2024).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eAdewojo, A. A., Amzat, O. B. \u0026amp; Abiola, H. S. AI-powered libraries: enhancing user experience and efficiency in Nigerian knowledge repositories. \u003cem\u003eLibr. Hi Tech. News\u003c/em\u003e. \u003cb\u003e42\u003c/b\u003e, 12\u0026ndash;16 (2025).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eChen, S., Xue, Y. \u0026amp; Cui, X. Information literacy of college students from library education in smart classrooms: Based on big data exploring data mining patterns using Apriori algorithm. \u003cem\u003eSoft. Comput.\u003c/em\u003e \u003cb\u003e28\u003c/b\u003e, 3571\u0026ndash;3589 (2024).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eNg, D. T. K., Leung, J. K. L., Chu, S. K. W. \u0026amp; Qiao, M. S. Conceptualizing AI literacy: An exploratory review. \u003cem\u003eComputers Education: Artif. Intell.\u003c/em\u003e \u003cb\u003e2\u003c/b\u003e, 100041 (2021).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eKorobenko, D., Nikiforova, A. \u0026amp; Sharma, R. Towards a Privacy and Security-Aware Framework for Ethical AI: Guiding the Development and Assessment of AI Systems. in \u003cem\u003eProceedings of the 25th Annual International Conference on Digital Government Research\u003c/em\u003e 740\u0026ndash;753Association for Computing Machinery, New York, NY, USA, (2024). \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1145/3657054.3657141\u003c/span\u003e\u003cspan address=\"10.1145/3657054.3657141\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eUnited Nations Educational, Scientific and Cultural Organization. Target 4.7: Sustainable development and global citizenship. (2019). Available at: \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://gem-report-2019.unesco.org/chapter/monitoring-progress-in-sdg-4/sustainable-development-and-global-citizenship-target-4-7/\u003c/span\u003e\u003cspan address=\"https://gem-report-2019.unesco.org/chapter/monitoring-progress-in-sdg-4/sustainable-development-and-global-citizenship-target-4-7/\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eVettriselvan, R. Empowering Digital Education: The Future of Value-Based Learning in the Digital Era. in \u003cem\u003eEmpowering Value Co-Creation in the Digital Era\u003c/em\u003e 199\u0026ndash;228IGI Global Scientific Publishing, (2025). \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.4018/979-8-3373-1742-7.ch009\u003c/span\u003e\u003cspan address=\"10.4018/979-8-3373-1742-7.ch009\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eAlmaleh, A., Aslam, M. A., Saeedi, K. \u0026amp; Aljohani, N. R. Align My Curriculum: A Framework to Bridge the Gap between Acquired University Curriculum and Required Market Skills. \u003cem\u003eSustainability\u003c/em\u003e 11, (2019).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eRajalakshmi, S. A. \u0026amp; Dynamic, L. J. N. R., P, V. Paul. NLP Enabled Chatbot for Rural Health Care in India. in \u003cem\u003eSecond International Conference on Computer Science, Engineering and Applications (ICCSEA)\u003c/em\u003e 1\u0026ndash;6 (2022). 1\u0026ndash;6 (2022). (2022). \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1109/ICCSEA54677.2022.9936389\u003c/span\u003e\u003cspan address=\"10.1109/ICCSEA54677.2022.9936389\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eAlqahtani, T. et al. The emergent role of artificial intelligence, natural learning processing, and large language models in higher education and research. \u003cem\u003eRes. Social Administrative Pharm.\u003c/em\u003e \u003cb\u003e19\u003c/b\u003e, 1236\u0026ndash;1242 (2023).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eAgostini, D. \u0026amp; Picasso, F. Large language models for sustainable assessment and feedback in higher education: Towards a Pedagogical and Technological Framework. \u003cem\u003eIntelligenza Artificiale\u003c/em\u003e. \u003cb\u003e18\u003c/b\u003e, 121\u0026ndash;138 (2024).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMbakwe, A. B., Lourentzou, I., Celi, L. A., Mechanic, O. J. \u0026amp; Dagan, A. ChatGPT passing USMLE shines a spotlight on the flaws of medical education. \u003cem\u003ePLOS Digit. Health\u003c/em\u003e. \u003cb\u003e2\u003c/b\u003e, e0000205 (2023).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHuang, H. et al. ChatGPT for shaping the future of dentistry: the potential of multi-modal large language model. \u003cem\u003eInt. J. Oral Sci.\u003c/em\u003e \u003cb\u003e15\u003c/b\u003e, 29 (2023).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eAl-Dokhny, A. et al. Can Multimodal Large Language Models Enhance Performance Benefits Among Higher Education Students? An Investigation Based on the Task\u0026ndash;Technology Fit Theory and the Artificial Intelligence Device Use Acceptance Model. \u003cem\u003eSustainability\u003c/em\u003e \u003cb\u003e16\u003c/b\u003e, 10780 (2024).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eAppleton, L. AI and academic libraries: what\u0026rsquo;s all the fuss about? \u003cem\u003eNew. Rev. Acad. Librariansh.\u003c/em\u003e \u003cb\u003e30\u003c/b\u003e, 1\u0026ndash;3 (2024).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eFu, Y., Weng, Z., Wang, J. \u0026amp; Examining AI Use in Educational Contexts: A Scoping Meta-Review and Bibliometric Analysis. \u003cem\u003eInt. J. Artif. Intell. Educ.\u003c/em\u003e \u003cb\u003e35\u003c/b\u003e, 1388\u0026ndash;1444 (2024).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLo, L. S. An initial interpretation of the U.S. Department of Education's AI report: Implications and recommendations for academic libraries. \u003cem\u003eJ. Acad. Librariansh.\u003c/em\u003e \u003cb\u003e49\u003c/b\u003e (5), 102761 (2023).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eEtikan, I. Sampling and Sampling Methods. \u003cem\u003eBBIJ\u003c/em\u003e 5, (2017).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBouncken, R. B., Czakon, W. \u0026amp; Schmitt, F. Purposeful sampling and saturation in qualitative research methodologies: recommendations and review. \u003cem\u003eRev Manag Sci\u003c/em\u003e (2025).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eJia, J., Yuan, Z., Pan, J., McNamara, P. E. \u0026amp; Chen, D. Decision-making behavior evaluation framework for LLMs under uncertain context. \u003cem\u003eProceedings of the 38th International Conference on Neural Information Processing Systems\u003c/em\u003e (NIPS '24) 37, Article 3601, pp. 113360\u0026ndash;113382 (2025).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eGrande, V., Kiesler, N. \u0026amp; Francisco, R. M. A. Student Perspectives on Using a Large Language Model (LLM) for an Assignment on Professional Ethics. in \u003cem\u003eProceedings of the 2024 on Innovation and Technology in Computer Science Education V. 1\u003c/em\u003e 478\u0026ndash;484ACM, Milan Italy (2024). \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1145/3649217.3653624\u003c/span\u003e\u003cspan address=\"10.1145/3649217.3653624\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eKozov, V., Ivanova, B., Shoylekova, K. \u0026amp; Andreeva, M. Analyzing the Impact of a Structured LLM Workshop in Different Education Levels. \u003cem\u003eAppl. Sci.\u003c/em\u003e \u003cb\u003e14\u003c/b\u003e, 6280 (2024).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eYadav, R. \u0026amp; Pathak, G. S. Determinants of Consumers\u0026rsquo; Green Purchase Behavior in a Developing Nation: Applying and Extending the Theory of Planned Behavior. \u003cem\u003eEcol. Econ.\u003c/em\u003e \u003cb\u003e134\u003c/b\u003e, 114\u0026ndash;122 (2017).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eChen, M. F. \u0026amp; Tung, P. J. Developing an extended Theory of Planned Behavior model to predict consumers\u0026rsquo; intention to visit green hotels. \u003cem\u003eInt. J. Hospitality Manage.\u003c/em\u003e \u003cb\u003e36\u003c/b\u003e, 221\u0026ndash;230 (2014).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eChen, Y. C., Ho, C. C. \u0026amp; Kuo, S. M. Service Quality of and User Satisfaction with Non-State-Owned Academic Libraries in China: Integrating the Fuzzy Delphi Method with the Kano Approach. \u003cem\u003eSustainability\u003c/em\u003e \u003cb\u003e14\u003c/b\u003e, 8506 (2022).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eEuropean Commission. Digital Education Action Plan. (2021\u0026ndash;2027). Available at: \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://education.ec.europa.eu/focus-topics/digital-education/action-plan\u003c/span\u003e\u003cspan address=\"https://education.ec.europa.eu/focus-topics/digital-education/action-plan\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eU.S. Department of Education, Office of Educational Technology. Artificial Intelligence and the Future of Teaching and Learning: Insights and Recommendations. (2023). Available at: \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://event-cdn.baai.ac.cn/file/file-browser/hAZcnhiJsJaGdyBnFhZjw3Wa7fPpGZ5D.pdf\u003c/span\u003e\u003cspan address=\"https://event-cdn.baai.ac.cn/file/file-browser/hAZcnhiJsJaGdyBnFhZjw3Wa7fPpGZ5D.pdf\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\u003c/ol\u003e"}],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":true,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":false,"hideJournal":false,"highlight":"","institution":"","isAcceptedByJournal":false,"isAuthorSuppliedPdf":false,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":false,"isPdf":false,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"scientific-reports","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":false,"externalIdentity":"scirep","sideBox":"Learn more about [Scientific Reports](http://www.nature.com/srep/)","snPcode":"","submissionUrl":"","title":"Scientific Reports","twitterHandle":"","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"stoa","reportingPortfolio":"Scientific Reports","inReviewEnabled":true,"inReviewRevisionsEnabled":true},"keywords":"Large-Scale Artificial Intelligence Models, Library Information Retrieval System, Education for Sustainable Development, Sustainable Development, Industry5.0","lastPublishedDoi":"10.21203/rs.3.rs-8751190/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-8751190/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003cp\u003eIn an era where the international community places significant emphasis on sustainable development goals, the strategic importance of education for sustainable development (ESD) is particularly pronounced. In the Industry 5.0 era, a growing number of university libraries have adopted Large-Scale Artificial Intelligence Models. This study surveyed 315 students and employed a multiple regression model to investigate the mechanisms through which the application of Large-Scale Artificial Intelligence Models to library information retrieval systems impacts ESD. The findings indicate that the integration of Large-Scale Artificial Intelligence Models into library information retrieval systems can enhance ESD while also improving users' information assessment capability, digital literacy, and awareness of privacy and security. Concurrently, the aforementioned skills, literacy, and awareness of uses served as mediating factors. These results remained consistent after undergoing a series of robustness tests. This study provides technical support and practical paradigms for enhancing the inclusiveness, precision, and long-term impact of education for sustainable development, as well as for cultivating interdisciplinary talents with a global perspective and green development concepts.\u003c/p\u003e","manuscriptTitle":"Promoting Education for Sustainable Development through AI-Enhanced Library Information Retrieval Systems","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2026-02-17 07:56:22","doi":"10.21203/rs.3.rs-8751190/v1","editorialEvents":[{"type":"communityComments","content":0},{"type":"decision","content":"Revision requested","date":"2026-03-23T10:14:25+00:00","index":"","fulltext":""},{"type":"editorInvitedReview","content":"","date":"2026-03-07T04:33:15+00:00","index":"hide","fulltext":""},{"type":"editorInvitedReview","content":"","date":"2026-02-28T05:44:47+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"330569763135835952145709687838891346012","date":"2026-02-17T09:03:14+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"339997643273418730075622366018966538084","date":"2026-02-14T08:59:35+00:00","index":"hide","fulltext":""},{"type":"editorInvitedReview","content":"","date":"2026-02-14T05:31:34+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"258902782713695720300454040510684810467","date":"2026-02-14T05:24:42+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"15319563413948972101747697800539446192","date":"2026-02-13T13:58:56+00:00","index":"hide","fulltext":""},{"type":"editorInvitedReview","content":"","date":"2026-02-13T09:10:02+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"84214371540932997560189074252618828799","date":"2026-02-12T09:40:38+00:00","index":"hide","fulltext":""},{"type":"reviewersInvited","content":"","date":"2026-02-12T08:41:44+00:00","index":"","fulltext":""},{"type":"editorAssigned","content":"","date":"2026-02-10T11:17:57+00:00","index":"","fulltext":""},{"type":"editorInvited","content":"","date":"2026-02-06T22:39:00+00:00","index":"","fulltext":""},{"type":"checksComplete","content":"","date":"2026-02-05T06:46:41+00:00","index":"","fulltext":""},{"type":"submitted","content":"Scientific Reports","date":"2026-02-05T06:36:26+00:00","index":"","fulltext":""}],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"scientific-reports","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":false,"externalIdentity":"scirep","sideBox":"Learn more about [Scientific Reports](http://www.nature.com/srep/)","snPcode":"","submissionUrl":"","title":"Scientific Reports","twitterHandle":"","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"stoa","reportingPortfolio":"Scientific Reports","inReviewEnabled":true,"inReviewRevisionsEnabled":true}}],"origin":"","ownerIdentity":"0a9ae317-b001-4879-a8a5-9ec95cf2bb70","owner":[],"postedDate":"February 17th, 2026","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"under-review","subjectAreas":[{"id":62924982,"name":"Social science/Education"},{"id":62924983,"name":"Business and commerce/Information systems and information technology"},{"id":62924984,"name":"Social science/Science technology and society"}],"tags":[],"updatedAt":"2026-04-09T21:38:16+00:00","versionOfRecord":[],"versionCreatedAt":"2026-02-17 07:56:22","video":"","vorDoi":"","vorDoiUrl":"","workflowStages":[]},"version":"v1","identity":"rs-8751190","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-8751190","identity":"rs-8751190","version":["v1"]},"buildId":"XKTyCvWXoU3ODBz1xrDgd","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}