The article analyzes the intelligent grading module based on generative AI developed by Tsinghua University Library for the course of "literature retrieval and utilization", introduces its construction ideas, architecture, and implementation process, selects literature research reports from students of different discipline background and grades as samples, and evaluates the application effect of the intelligent grading module from three aspects: grading consistency, correction quality and efficiency, and student feedback and acceptance. The results indicate that scores of manual grading and intelligent grading are significantly consistent; the intelligent grading can significantly improve the efficiency of literature research report grading and the quality of information feedback; students' acceptance of intelligent grading is high, but they prefer the "humanAI collaboration" mode that combines intelligent and manual grading. Based on this, relevant suggestions are proposed, namely construct a course adaptation and dynamic evaluation system, explore the construction of localized models and highquality data feeding, strengthen data privacy protection and algorithm ethics supervision.

The article summarizes research status of future learning centers in university libraries, by exploring the collaborative relationships between data, platform, and space, it explains logical mechanism of future learning centers construction in university libraries, clarities construction framework and principles of future learning centers in university libraries based on the "trinity" of data, platform, and space, proposes two specific practical paths of "gradual" and "leapfrog": namely, build a steady development model that includes integrating multimodal data, optimizing and upgrading digital service platforms, and reconstructing and expanding virtual and real spaces, as well as an innovative development model that includes establishing precise learning service ecosystem, creating globally interconnected intelligent hub, and creating multidimensional collaborative learning scenarios.

The article discusses the supportive role of generative artificial intelligence in public library reading therapy services, focusing on three dimensions: building user profiles, strengthening resource support, and promoting psychological interaction. It designs public library reading therapy service concepts based on generative artificial intelligence and proposes corresponding service strategies, namely comprehensively extract user demand tags and customize reading therapy service plans; customize and design a reading therapy book to creat reading therapy space; create a longterm interactive loop for users and guide them to shape healthy psychology.

The article reviews the data subject rights norms in China's public cultural service data empowerment domain related policies and regulations, and finds that the existing rights structure in the field of public cultural service data empowerment suffers from the virtual protection of data source copyright, centralized operation management rights of data processors, and the digital divide of basic cultural rights of data users. And based on the strong public nature of public cultural service data and the functional positioning of its public goods, this paper proposes implementation strategies for the reform of rights separation in public cultural service data, namely constructing triple rights system, which takes the copyright of the data source as the primary right, the business rights held by the data processor as the derivative right, and the basic cultural rights of the data user as the purposive right.

This paper analyzes the connotation characteristics of the full smart library, including multimodal resource presentation, full scene smart service, virtual and real space interaction, immersive sensory experience, collaborative innovation of swarm intelligence and green digital ecology. Based on the five core elements of environment, technology, resources, service and management, this paper puts forward the construction strategies of the full smart library, that is, create immersive holographic digital environment, increase the enabling intensity of metaverse technology, expand the resource acquisition channels of virtual and real interaction, improve the intelligent service ability of the whole scene, and build a highly integrated comprehensive management system.

This article elaborates on the connotation of tip literacy, difference and connection with tip project and AI literacy, systematically sorts out the core elements of the cultivation goals and framework of tip literacy in university libraries. Aiming at the existing problems in the cultivation of tip literacy in university libraries, it proposes optimization strategies, namely constructing a threedimensional cooperation network for the cultivation of tip literacy, comprehensively build a team of tip literacy cultivation teachers, systematically create tip literacy cultivation resource system, and personalized design implementation modes for tip literacy cultivation.

The article analyzes the constituent elements of traditional culture promotion in university libraries based on the theory of interactive ritual chain, and constructs implementation framework for traditional culture promotion in university libraries with the core of building interactive scenarios, stimulating users' deep participation willingness, and disseminating traditional cultural values, and proposes practical paths for the traditional culture promotion in university libraries, namely create diverse contexts and shape interactive rituals; enhance emotional resonance and deepen participation motivation; expand the scope of promotion and promote value appreciation.

Based on sorting out and analyzing the relevant research on the subject services evaluation in university libraries, this article uses the network analytic hierarchy process to construct index system for the subject services effectiveness evaluation of university libraries, which includes 4 primary indicators, 9 secondary indicators, and 37 tertiary indicators. To verify the feasibility and effectiveness of the indicator system, this paper selects a university library in Tianjin for empirical evaluation. Based on the fuzzy comprehensive evaluation method, it calculats subject service effectiveness score of the library, and based on the analysis results, proposes suggestions for the subject service of university library, namely intensify the promotion of digital resources and leverage the advantages of disciplinary resources; standardize the system of subject librarians and clarify the paths of subject service work; introduce AI tools to assist with disciplinary service work.

The article focuses on the structural conflicts between technological efficiency and humanistic values in the transformation of library services in digitalintelligent era, and proposes the concept of "human+AI" dual librarians service symbiosis. By clarifying the service symbiosis goals of "human+AI" dual librarians' complementary functions, collaborative efficiency, and ecological reconstruction, this paper systematically analyzes the service symbiosis mechanism of "human+AI" dual librarians including the virtual real linkage of intelligent services, knowledge service resonance of value isomorphism, and emotional connection and integration of humanistic enhancement. It proposes three in one "human+AI" dual librarians service symbiosis paths of institutional innovation, technological empowerment, and ecological cultivation.

On the basis of explaining the connotation of data governance in smart library under the background of generative artificial intelligence, this paper analyses the composition of data governance system of smart library under the background of generative artificial intelligence from five dimensions of the top-level design, support, implementation guidance, practical application and bottomlevel planning, focusing on the objectives, tools, subjects, processes and objects of data governance, and points out that the data governance of smart library faces the realistic dilemma of the value deviation of governance objectives, the implementation deviation of governance tools, the imbalance of governance subject structure, the improper standardization of governance process, and the insufficient validity of governance objects. Then from five perspectives of value guidance, tool integration, subject collaboration, process specification, and object quality improvement, this paper puts forward the transformation paths of data governance in smart library under the background of generative artificial intelligence, namely reshaping the value orientation and hierarchy of data governance, promoting the management adaptation of systems tools and technical tools, forming a diversified ecological pattern of data governance, carrying out the whole process of data security compliance governance, and strengthening the refined control of basic data quality.

The article is based on the protective motivation theory and perceived value theory, and constructs a configuration model on the influencing factors of the privacy protection disengagement for social media users. The NCA and fsQCA methods are used to analyze the influencing factors of the privacy protection disengagement and their configuration paths. Research has found that the interaction between threat, response assessment, and value assessment form four types of configurations that can lead to high privacy protection disengagement, namely emotional value dominant, response cost driven, response efficacy deficiency and functional value dominant, and response cost and social value interaction driven. Among them, emotional value plays a universal role, and privacy concerns and response costs are crucial in suppressing the disengagement of privacy protection. Suggestions have been put forward to guide users' privacy protection behavior.

This paper uses the experimental method to study 110 undergraduate students of the same grade and major in a university in Tianjin, randomly divides them into three groups, and carries out the health information literacy intervention by using picture, text and video modal educational resources, and measures intervention effect and analyzes by pretest and posttest questionnaires. The results show that: the overall level of university students' health information literacy is low; three kinds of modal educational resources can effectively improve their health information literacy, but there are significant differences in the improvement effect, among which the image education resources has the most significant improvement effect; the promotion effect of educational resources on university students' health information literacy is affected by a variety of factors including annual family income level and parents' education level. Based on this, this paper puts forward suggestions for improving university students' health information literacy, namely, constructing systematic education intervention system, developing differentiated education resources, and designing personalized education programs.

The article analyzes the concept characteristics, logic reasoning, and realization paths of datadriven think tank knowledge service innovation, points that data-driven think tank knowledge service innovation is realized through the process mechanism of data collection, processing, storage, calculation and application through demand perception, data fusion and knowledge transformation into content architecture, and three paths of service datalization, data servicalization and recycling. Accordingly, it proposes innovation development strategies, namely deepening the service concept and expanding the service content; deeply cultivating digital technology and optimizing the service model; gathering resource elements and consolidating data capabilities.

The article selects 87 universities among the top 100 universities in the QS world university rankings that have already offered artificial intelligence (AI) literacy courses as the research object. It investigates and analyzes the construction of AI literacy courses in the sample universities from four dimensions: course leading knowledge, course teaching plan, course content, and course assessment method, summarizes the characteristics. Based on this, inspirations for AI literacy courses construction in Chinese universities are proposed, namely, designing course content based on the KSEAE model; combining theory with practice, emphasizing practical teaching; promoting the construction of "AI+" general courses; strengthening AI ethics and social responsibility education; strengthening international exchange and cooperation, and introducing international highquality courses resources.

Based on the practice of the classic ancient books reading promoting in Guangzhou Library, this article constructs classic ancient books reading promoting model in public library including three dimensions: cultural translation, experiential reading, and memory reconstruction, and proposes practice strategies for the classic ancient books reading promoting in public library, namely optimizing content translation and reducing comprehension barriers of users; innovating experience design and enhancing user interest and engagement; advancing memory reconstruction, enhancing cultural memory and identity; improving guarantee mechanism and enhancing promotion efficiency of ancient books.

The article explains the connotation and characteristics of AI literacy, analyses the role of reading promotion in cultivating AI literacy, and constructs a AI literacy cultivation plan based on reading promotion of university libraries according to the ARCS model, which covers attracting user attention, linking user needs, designing support systems, and enriching service scenarios. It also proposes cultivating paths, namely enhance reading willingness of users, and achieve selfdirected learning; break down barriers to business integration and promote functional connectivity; refactor the literacy of librarians and improve AI literacy cultivation support services; create diverse contexts, establish incentive mechanisms and evaluation systems for cultivating AI literacy.

Based on UTAUT2 model, this paper constructs a user dialogue based search behavior influencing factor model based on four variables: performance expectations, effort expectations, convenience conditions, and price value. Validate the path relationships between variables through structural equation modeling and analyze the importance of each variable using artificial neural networks. The results indicate that 4 variables all have a significant positive impact on user conversational search behavior, with performance expectations and effort expectations as internal driving factors having the most significant impact on user conversational search behavior, followed by convenience conditions and price value as external driving factors. The empirical results of ANN on the importance of independent variables are consistent with the SEM conclusion, further enhancing the reliability of the SEM conclusion. On this basis, this article proposes strategies for improving the user experience of conversational search platforms, which is to build a content output verification mechanism and strengthen user trust in technical performance; build a layered application interaction system to reduce the burden of users' technical cognition; establish an intelligent embedded service model to enhance the convenience of user technology usage; implement a gradient pricing strategy to optimize users' perception of technology costs.

This article analyses the existing problems of regional joint collection construction in public libraries, designs an framework for regional joint collection construction in public libraries covering the data layer, facility layer, document layer and institutional layer based on the application advantages of intelligent stereoscopic book depository, and proposes specific practical strategies from the empowerment perspective of intelligent stereoscopic book depository, that is, efficiently aggregate regional collection resources by using intelligent stereoscopic book depository, improve the regional joint collection environment with the help of intelligent stereoscopic book depository, dynamically optimize the regional joint collection structure relying on intelligent stereoscopic book depository, and enhance the management of regional joint collections through intelligent stereoscopic book depository.

The article constructs a research model on the influencing factors of user transfer behavior in large language model based on the PPM theory, and proposes corresponding path hypotheses. The empirically tested and configurational analysis is conducted of model hypotheses. The results show that information quality and usage habits have a significant positive impact on user satisfaction, satisfaction has a significant negative impact on user transfer behavior; relative advantage, technological availability, and usage context have a significant positive impact on users' perceived value, and perceived value has a significant positive impact on users' transfer behavior; the significant positive impact of social influence on user transfer behavior. Based on this, the fsQCA method is used to identify three configuration paths, and it explores directions for improving and optimizing large language models around core conditions such as usage habits, relative advantages, usage contexts, and social influence.

The article takes data literacy practical education of representative 10 university libraries in China as examples, and uses literature research, online research, and case analysis methods to systematically analyse the current situation and problems of data literacy practical education in Chinese university libraries. It also proposes development strategies for data literacy practical education in university libraries, namely strengthen multiparty collaborative cooperation to achieve resource coconstruction and sharing; promote the diversified supply of practical education carriers， innovate practical education forms; optimize practical education content and enhance user data literacy; make good use of incentive measures, stimulate users data innovation potential; improve multidimensional dynamic evaluation, and build intelligent feedback mechanisms.

This article explores the evaluation elements of digital resources inclusive services in university libraries based on grounded theory. It constructs models for evaluating digital resources inclusive services in university libraries from five dimensions: identifying service objects, optimizing resource supply, improving service adaptability, constructing service frameworks, and strengthening evaluation feedback. The specific connotations of the five evaluation dimensions and their internal logical mechanisms for promoting the improvement of digital resources inclusive services in university libraries are explained, and research implications are proposed, namely, integrating the evaluation subject and constructing the evaluation community with multiple participation; developing evaluation tools and building an intelligent collaborative technical support system; optimizing the evaluation mechanism and forming the operation process of dynamic management.

From the perspective of bidirectional empowerment, this paper systematically deconstructs the internal coupling mechanism between the threedimensional core of technology, factors, and industry of new quality productivity and functional paradigm transformation, element recombination and upgrading and ecological niche restructure of future learning centers. It constructs a threedimensional collaborative mechanism encompassing technology, organization, and institution, and proposes phased implementation paths oriented toward foundation reconstruction, efficiency enhancement, and ecological leapfrogging, namely collaboratively reconstruct technology base and spatial functions, collaboratively promote organizational capacity upgrading and system innovation pilot, collaboratively construct digital twins, integration of industry and education and regional innovation ecology.

Using online survey methods, this paper investigates the current situation of open education resource (OER) construction and service in Finnish university libraries from three aspects: resource search and sharing, evaluation and adoption, creation and publication, summarizes the characteristics of OER, and proposes inspirations for promoting OER development in China, including improving the OER supportive policy system, carrying out full process OER services, cooperating on the development and layout of OER platforms, enhancing librarians' OER service capabilities, and exploring mechanisms of direct highquality OER to the grassroots.

The article constructs a theoretical framework of "demand identification—strategy matching", combines with a health promotion ecological model and demand side data analysis, analyzes university students' health service needs in dimensions of individual, interpersonal organization, community and policy, clarifies intervention strategy priorities, and then proposes health promotion strategies for university libraries, namely deepening health knowledge services at the individual level; building supportive social networks and leveraging peer education effectiveness at the interpersonal level; integrating the campus health resource network at the community level; optimizing health service environment of library at the organizational level; promoting institutional health support at the policy level.

This paper analyzes the emotional education value of emotional design of library digital resources, such as driving users' independent development, conserving users' cultural foundation, and promoting users' social participation, constructs the emotional design model of library digital resources, and puts forward the implementation strategies, namely clarify the cultivation goal of emotional civilization and promote users' emotional focus; systematically develop emotional resources and build an open and stable resource generation mechanism; build immersive scenes based on user needs to achieve accurate emotional guidance; carry out practical tests and constantly improve the emotional design model.

Taking digital technology driven as core perspective, this paper systematically analyzes the core connotation and main characteristics of trusted data space in smart library driven by digital technology, explains the construction logic of trusted data space in smart library driven by digital technology from two dimensions of "technology -system" collaborative evolution and dynamic interaction among multiple subjects, analyzes the realization mechanisms of the trusted data space in smart library driven by digital technology based on three stages of data element reconstruction, governance process reengineering and service value transition, and puts forward specific construction paths based on three levels of resources, capabilities and applications, and then puts forward corresponding guarantee strategies from four levels of institutional norms, technical support, talent cultivation and ecological collaboration.

This paper analyses concept of new cultural spaces and smart reading service ecosystem of university library, introduces the theory of empowerment, explains enabling mechanism and theory framework of smart reading service ecosystem construction empowered by new cultural space of university library, and puts forward the strategies of smart reading service ecosystem construction of university library, namely deepen the space reconstruction and creat education oriented compound reading space; pay attention to knowledge driven and build knowledge service system of global integration; embed digital intelligence technology and create smart reading scene; strengthen crossborder cooperation and expand social reading service network.

This paper builds a theoretical model based on the influencing factors of college students' approachavoidance behavior towards artificial intelligence generated content (AIGC) in the grounded theory of academic environment, uses structural equation modeling to investigate the determinants of AIGC approachavoidance behavior in academic environment. Research results show that psychological empowerment and relative advantage have a significantly positive impact on emotional attachment; AI illusion and privacy concern significantly positively influence perceived eeriness; emotional attachment and individual innovativeness significantly positively affect approach behavior; perceived eeriness significantly positively influence avoidance behavior; and academic ethics play a negative moderating role between emotional attachmentapproach behavior. Based on this, the article puts forward optimization suggestions of college students using AIGC of academic environment: optimizing platform functions, enhancing emotional attachment of college students; improving AI literacy of college students, strengthening information security barriers; building multi-stakeholder support system, jointly promoting academic integrity.

This paper constructs an behavior strategic evolutionary game model of four agents: users, public libraries, smart service platform and government, and conducts simulation, explores the equilibrium stable of the value co-creation system for smart services of public library under the agents' behavior strategic interactions, clarifies the logic of how key parameters of the model influence the evolution process of agents behavior strategy. Simulation results indicate that the government's support initial willingness plays a core leading role in the smart service value cocreation system. Through policy and financial support, it drives public libraries, smart service platforms, and users to evolve towards the behavior strategies of "highinvestment construction" "providing highquality service" and "active participation" respectively; incentives affects the enthusiasm for participation of users and investment construction level of public libraries; resource allocation ratio of platform incentivizes smart service platforms to improve service quality, may also constrain public library to evolve towards positive behavior strategies; government accountability strengthens the smart service platforms' sense of responsibility, prompting them to provide high-quality services. Based on this, it puts forward optimization strategies of smart service value cocreation in public libraries.

Taking the cultural tourism integration as the research background, this paper expounds the concept and characteristics of immersive exhibition, analyzes the current situation of library immersive exhibition service from four aspects: spatial form, experience type, theme content and audience group, points out that there are problems of imbalance between technology application and cultural content, serious homogenization and lack of innovation, cultural distortion and ethical disputes, and puts forward library immersive exhibition innovation paths from three innovative paths of focus on content activation and narrative reconstruction, innovate cultural expression forms, and strengthen interactive participation.

This study takes the SERVQUALIPA model as foundation, based on the user perspective constructs an evaluation index system of bibliotherapy services quality of university libraries consisting of 7 primary indicators and 29 secondary indicators, and assigns index weights using the analytic hierarchy process. Furthermore, an questionnaire survey and empirical study are conducted facing college students based on this evaluation system. The IPA quadrant diagram is adopted to account the service quality of bibliotherapy. Strategies for improving bibliotherapy service quality of university libraries in the AI era are accordingly proposed: empowering the prescription matching and interactive process of bibliotherapy services relying on the integration of AIGC and knowledge graphs, embeding successful practices of the AIgenerated dialogue system for bibliotherapy into team implementation procedures, activating empathic connections using AI enabled longshortterm memory and affective computing, and compensating for the deficiencies of solely relying on the AIgenerated dialogue system through offline physical bibliotherapy services.

This paper uses the method of purposive sampling and maximum difference sampling to select 32 university libraries as research samples from multiple channels. It systematically analyzes the status and existing problems of bibliotherapy service from three dimensions of service form, space construction and professional support, and puts forward development paths of bibliotherapy service in university libraries, namely deepen service connotation and innovate healing form; promote spatial integration and build collaborative ecology; enhance strategic positioning and improve cooperation mechanism; establish librarian training system and strengthen talent support; improve evaluation system and drive optimization cycle.

Based on elaborating on the value of the research topic and reviewing relevent research, the article introduces the overall framework, main content, research ideas and methods, difficulties, innovations, expected objectives, ralue and significance of the National Social Science Fund Major Project " Comprehensive Research on the Value Interpretation, Display and Utilization of Important Intangible Cultural Heritage Empowered by Digital Intelligence". This project aligns with the "creative inheritance and innovative development" requirements for China's outstanding traditional culture, focuses on the core of digital intelligence empowerment, and aims to foster the coordinated development of systematic protection and innovative inheritance of intangible cultural heritage, constructs comprehensive research framework of " historical tracing—value decoding—gene preservation—innovative revitalization—identity reinforcement". Through the coordinated advancement of five subprojects, providing theoretical support and practical solutions for the implementation of the national cultural digitalization strategy and the deep integration of culture and tourism.

 This paper analyses and points that achieving good data governance and date empowerment is the motivation of library data governance from three "metaquestions": why, what and how to govern, clears that its governance scope needs to extend from the traditional multisource heterogeneous "data body" management to the dynamic evolution of "association relationship", at the same time, the algorithm is included in the regulation scope. Based on this, the realization paths of library data governance are proposed, namely move towards good data governance by creating high-quality, interoperable, safe and compliant "data body"; promote data empowerment by build "relational" governance paradigm; control algorithm governance by build "responsible" intelligent data governance mechanism.

Based on online surveys, literature analysis, and semistructured interviews, this article extracts 26 demands for AI research support services from university library users, and applies the Kano model to investigate users' needs for AI research support services. Based on traditional Kano model classification, hybrid classification, and BetterWorse index classification analysis, user needs are classified into four attributes: basic, expected, attractive and indifference, and prioritized, furthermore, it proposes optimization strategies for research support service in university libraries under the AI4S background.

Taking DeepSeek as an example, this paper analyzes the technical mechanism of public library health information service innovation driven by large language model, and explains the evolution process of public library health information service paradigm driven by DeepSeek from four dimensions: demand perception, information production, service supply and value creation. Based on this, this paper constructs public library health information service scenario based on DeepSeek, which is composed of information push service, literacy education service, digital inclusion service, emergency response service and crossborder integration service, and puts forward public library health information service implementation paths driven by large language model from four aspects: technology docking, resource construction, librarian empowerment and mechanism guarantee.

This paper investigates and analyzes the quality control practice status of service outsourcing in 16 different levels public libraries in Fujian Province, points out problems such as insufficient quantification of indicators setting, qualitative distortion and lack of library level adaptation, incomplete supervision coverage and manpower limitation, shortage of funds and team instability, lack of standards and policy lag, and further puts forward optimization strategies for public library service outsourcing quality control, namely the construction of quantitative and qualitative double track assessment index system, the building of digital empowerment system and professional supervision team, the construction of funding and talent training system, and the construction of standard guidance and policy support system.

Taking the DNA double helix structure as a metaphor, this paper constructs a dualchain coupling model of questioning ability collaborative evolution between librarians and prompt engineers, and verifies the validity of the model through a comparative experiment. The results show that the collaborative group composed of librarians and prompt engineers is significantly better than the single professional team in the quality of completing complex knowledge tasks, which shows that the model can produce systematic advantages beyond the sum of individual abilities. On this basis, the collaborative improvement strategies of librarians' and prompt engineers' questioning ability are put forward, namely establish demandoriented crossborder learning and knowledge exchange platform, promote technology integration and questioning paradigm reconstruction based on skill transfer, deepen scene co construction and business innovation practice based on project driving, and focus on value co creation and build the ethical norms and collaborative governance system.

Based on the analysis of "structure—process—function" research framework, this paper points out that AI large model empowering the construction of new think tank in libraries by reconstructing the organizational form, changing the operation process, and deepening think tanks functions; and it defines the challenges it faced in the new think tank construction in library, such as organizational structure and governance mechanism, model training and application promotion, function customization and security guarantee, and puts forward specific practical paths, namely promote organizational structural change and build talent support system; speed up process reengineering and promote technology application and promotion; develop multiple functions and strengthen network security guarantee.

On the basis of reviewing relevant research status, this article elaborates on the mechanism of network security situation awareness based on open source intelligence, analyzes the technical empowerment mechanism of generative AI in this field, and constructs network security situation awareness model based on open source intelligence empowered by generative AI. It explores the operational bottlenecks faced by the model and proposes safeguard strategies, namely establishing high-quality training data system and false information governance system, optimizing computing resource allocation and reducing model deployment costs, and improving the corresponding ability of the algorithm, optimizing data processing and storage architecture.