Artificial intelligence, a machine learning technology that aims to enhance human intelligence, will undoubtedly have a profound impact on the development of human society. Museums, as institutions that document the historical processes of human and natural development, are tasked with preserving, processing, researching, and applying a vast amount of information. This article summarizes the characteristics of this information, analyzes the traditional models of museum informatization, and proposes methods to integrate and utilize artificial intelligence concepts and technologies in museum information. Artificial intelligence technology seeks to innovate traditional museum information systems by approximating human cognitive processes and methods of utilization, overcoming the shortcomings of traditional information systems in terms of information use, application expansion, and system maintenance, thereby making information processing technology a fundamental aspect of museum work.

The implementation of the national cultural digitization strategy and the promotion of the intelligent utilization of ancient books in the new era have put forward requirements for ancient book collection service institutions to use modern technological means to comprehensively improve the protection and dissemination capacity of ancient books. The technological development and exploration practice of the metaverse library provide a development path, theoretical support, and technical support for the collection of ancient books in the new era, and promote the reform and innovation of the work mode of ancient books. This paper reviews the development process and technical characteristics of the metaverse, and analyzes the theoretical basis, and value cognition of the metaverse to empower the work of library special collections，discusses the supporting role of metaverse technology in the construction and service of special collections of libraries from five aspects: exploring spatial reconstruction, reshaping the form of resources, providing immersive services, constructing a sharing mechanism, and cultivating professional talents, so as to make special collection resources “live”, special collection space “integrated”, and special collection services “moved”.

In the context of increasingly tense geopolitical situations and the rapid development of artificial intelligence technology, this paper discusses the construction and support of open access resources in Chinese universities, providing a reference for safeguarding national scientific and technological security, cultural security, information security, data security, etc. The paper analyzes the development process of open resources, as well as the current status of open resource construction and utilization in domestic universities. It points out the main problems of current construction open resources in universities in terms of security and confidentiality awareness, independent control of basic platforms, core technology innovation, and resource outflow. It proposes measures and suggestions for construction and support of open resources in universities, including organizational aspects, policy norms, technological innovation, new generation open resource platforms, resource review and evaluation, resource deposit mechanism, open cooperation, etc.

In recent years, while the construction of smart libraries in China has been vigorously promoted, it is also actively exploring new modes of intelligent services. Recently, large language models have attracted widespread attention from various industries due to their outstanding performance. With the open source of multiple large language models and the gradual maturity of the development paradigm of “Model as a Service (SaaS)”, smart libraries have ushered in the opportunity to upgrade their services with the help of library large language models. This paper outlines the basic concepts of large language models, proposes the basic ideas for constructing library large language models, explores the application directions of large models in the library industry, and analyzes the challenges faced by libraries in applying large language models.

The constantly changing demand of users for literature information is a normalized phenomenon. How libraries can determine the correlation between user needs and target literature transmission to improve library service efficiency is currently a problem worthy of research. Therefore, a development trend prediction method of library literature delivery services based on the LightGBM algorithm is proposed. After training the LightGBM model with the best hyperparameters,obtain the prediction development trend of library literature delivery services. The experimental results show that this method can effectively reduce the dimensionality of collected data, improve prediction efficiency, and achieve accurate prediction of the development trend of library literature delivery services.

This paper explores the coexistence and integration strategies of traditional MARC cataloging format and BIBFRAME bibliographic framework plan in the current domestic library environment. By analyzing the advantages and challenges of MARC and BIBFRAME, this paper indicates that although MARC format has strong description capabilities and high standardization, it faces the limitations of data closure and complex structure; while BIBFRAME,with its openness, semanticization, and efficient search capabilities, provides a new direction for library resource description.To address the difficulties in converting between the two formats, this paper proposes a coexistence strategy based on a core set of metadata elements, aiming to ensure the continuity of library resource description and promote book catalog data towards a more open, semantic, and efficient direction. By referring to successful international cases, such as the FOLIO platform architecture, this paper constructs a hybrid cataloging system, providing a practical path for the promotion and application of BIBFRAME in the domestic library field.

Large Language Models (LLM) is reshaping the technological landscape with a rapid momentum, and becoming a powerful engine for intelligent transformation in various domains. Driven by this technological wave, the library field has ushered in a historic opportunity for transformation.This article aims to explore the potential and possibility of the application of LLM in smart libraries. It analyzes the impact areas, impact methods, and application value of LLM in the library, summarizes the current characteristics of LLM implementation status in the library industry, and proposes application pathways and strategies. The study examines the potential AI application requirements in libraries from four aspects:services, operations, management, and space. Furthermore, it constructs and envisions the application scenarios of LLM in eight typical domains, including consulting, resource discovery, reading promotion, academic services, acquisitions and cataloging, resource processing, digital humanities, and management decision-making.

In recent years, the exhibition services of public libraries have developed rapidly, and the application of virtual exhibitions and XR technology in exhibitions has attracted increasing attention in the field of public libraries. XR virtual exhibitions in public libraries, considering the characteristics of their exhibits and their mission, should adopt more suitable technical choices and narrative logic. This paper takes the “Reading People, Reading Cities, Reading the World” headsetbased MR virtual exhibition project of the Shanghai Library as an example. From a practical perspective, it summarizes the project's practices in three areas: project positioning, technology, and content system. It explores the reconstruction of narrative logic in this XR virtual exhibition and considers the value of XR virtual exhibitions for public libraries. This includes the unification of the textual and historical value of documents, the expansion of traditional exhibitions, and the structured transmission of historical events. The paper also reflects on issues such as the limitations of technical platforms and the high entry threshold for entering the XR era, aiming to provide new ideas for the development and future practices of virtual exhibitions.

In the context of smart libraries, using intelligent methods to enhance the utilization rate of print literature and the efficiency of loan services is an important task for public libraries. By analyzing data of public libraries in Jiangsu Province from 2019 to 2023, the paper aims to provide reference strategies for improving the loan service efficiency of public libraries in China. The study found that public libraries in Jiangsu currently face challenges such as insufficient comprehensive service capabilities, concentrated service hours, documentation resources, and service group. It is recommended that public libraries can actively apply technologies such as big data and Artificial Intelligence,optimize service time and accessibility, strengthen institutional collaboration, and implement targeted reading promotion strategies to enhance the overall effectiveness of loan services.

Based on the user’s perspective, this paper explores the differences between the large language model search engine and the traditional search engine in information retrieval behavior and effectiveness, aiming to provide guidance and suggestions for the design and optimization of large language model in text content generation and return mechanism, so as to improve user experience and retrieval efficiency. Combined with questionnaire survey and observation method, this paper collects users' behavior data when using large language model search engine and traditional search engine for different retrieval tasks. Using screen recording and post-task questionnaire survey methods,records and analyzes user retrieval behavior and effectiveness indicators, and uses non-parametric test analysis to compare the differences in user retrieval behavior under different search engine environments. The research shows that compared with the traditional search engine, the large language model search engine has a long retrieval structure and less reconstruction times, the response is slow but the retrieval efficiency is high, and the result quality evaluation is insufficient but the user satisfaction is high. The research results provide directional guidance for the future evelopment and innovation path of large language model search engines,which are of great significance for promoting the continuous technological progress.

With the rapid development of artificial intelligence technology, the application of large language models in the field of library has gradually attracted attention, but there is little research on the use of large language models in library acquisition and cataloging business. This paper first introduces the main process of the acquisition and cataloging work in Shanghai library,and then analyzes the ability of large language models and the application prospects of large language models in the acquisition and cataloging business work . Through scenario examples,this paper explores the specific applications of large language models in five aspects such as the intelligent pre-selection of book lists, analysis of reader behavior and automated cataloging, and proposes a preliminary idea of integrating these applications into the library system.

During the Republican era in China, libraries and museums maintained a close and intertwined relationship, shaped by both foreign influences and indigenous factors. This article adopts a historical research approach to trace the transnational trajectory of the concept of “library-museum integration”—originating in the United Kingdom, transplanted to Japan, and subsequently introduced to China, where it underwent localized transformation. Under the influence of various educational ideologies including auxiliary education, social education, vocational training, and new democratic cultural education—the concept of “library-museum integration” in China manifested in several distinctive forms, including dual-purpose institutions, parallel development, scholarly community, and interdisciplinary academic programs. The article argues that institutional inertia rooted in aristocratic traditions played a key role in shaping the British model of “library-museum integration”. In Japan, driven by industrial development, the concept was functionally restructured. In the Chinese context, educational factors proved decisive in the evolution of the concept of “library-museum integration” in modern China.

Recently, DeepSeek has emerged as a phenomenal product due to its high efficiency, open-source nature, and low-cost training, fueling a nationwide AI boom. This interview explores the profound impact of the large AI model DeepSeek on the library industry. During the discussion, Mr. Liu Wei elaborated on DeepSeek’s technical advantages and analyzed its potential in library services such as knowledge provision, reference consultations, and intelligent agent applications. Mr. Liu emphasized that libraries must transform to adapt to the AI era, suggesting measures such as deploying local models, developing intelligent agents, and training readers to enhance library services. Furthermore, he pointed out that as technological and knowledge gaps continue to widen, librarians should actively embrace the changes of the times and technological advancements while steadfastly upholding their role as knowledge intermediaries.

This paper proposes a concept for systematically managing museum data using artificial intelligence technology based on attention mechanisms and explores the technical implementation methods. Specifically, it involves integrating all museum data into a unified data warehouse and employing large language models to interpret requests expressed in natural language, translating them into programming languages capable of processing multimodal data such as images, text, and audio. These languages are then used to perform tasks such as statistical analysis, querying, summarization, and reasoning. Through the collaboration of the data layer, processing layer, and interaction layer, the system not only ensures the accuracy, completeness, and comprehensiveness of the extracted data but also derives information from data analysis, integrates information into knowledge, and ultimately develops intelligent capabilities. Additionally, it addresses the long-standing issue of data silos within various museum business systems.

This paper proposes a concept for systematically managing museum data using artificial intelligence technology based on attention mechanisms and explores the technical implementation methods. Specifically, it involves integrating all museum data into a unified data warehouse and employing large language models to interpret requests expressed in natural language, translating them into programming languages capable of processing multimodal data such as images, text, and audio. These languages are then used to perform tasks such as statistical analysis, querying, summarization, and reasoning.Through the collaboration of the data layer, processing layer, and interaction layer, the system not only ensures the accuracy,completeness, and comprehensiveness of the extracted data but also derives information from data analysis, integrates information into knowledge, and ultimately develops intelligent capabilities.Additionally, it addresses the long-standing issue of data silos within various museum business systems.

Social character, as a product of the interaction between individuals and society, encompasses three dimensions:material self, social self, and spiritual self, reflecting the individual’s life priorities and value pursuits. This study takes the 2019 library borrowing data across various districts of Shanghai as a lens to examine the reading preferences of local residents and their corresponding social character traits, as well as the interactions between these traits and also explores the influence of dimensions such as region and age on social character. The BERTopic model is employed for topic clustering,and two-way ANOVA is applied to conduct a statistical analysis of the social character traits across different districts and age groups. It was found that the stage of regional development is closely related to social character traits. Residents in central districts tend to prioritize the spiritual self, whereas those in suburban or economically developed areas place more emphasis on the material self and social self. Age also has a significant impact on individual social character, with adolescents,adults, and elderly groups showing greater concern for the spiritual self, social self, and material self, respectively. Furthermore, residents’ social character is also influenced by the interaction between age and regional factors.

We are entering an era where technology is developing at an exponential rate,and it is anticipated that the “Singularity” will arrive within the next two decades. Large language model technologies have paved possible paths for the arrival of the “Singularity”. As humanity steps into the era of “Homo Technicus”, where humans and technology deeply integrate and co-evolve, traditional humanism centered on humans faces challenges: new humans may significantly alter their nature through technological integration; machine intelligence, through self-acceleration and iteration, might develop autonomous consciousness; and the development of AI brings forth various risks and ethical issues, even concerning survival or destruction. Immediate action is necessary to incorporate the issues brought about by AI development into the research scope of the new humanities. In the next twenty years, digital humanities need to make a new turn, following the expanded new humanism, and utilize the fifth paradigm of digital humanities research (AI4DH) to evolve towards wisdom humanities.

In the context of the era of Digital Intelligence and Humanities, the paper discusses in depth the challenges encountered by the humanities, the integration and development of GLAMS and their mission of the times. Driven by the scientific and technological revolution, the integration of GLAMS has become a trend, and they show competition and integration in the public cultural service sector. The mutual integration of GLAMS not only promotes the dissemination and acquisition of knowledge, but also gives rise to the emergence and development of new knowledge. Modern information technology provides GLAMS with an eternal space that transcends physical boundaries and material life, so that the information and values they carry can be disseminated and passed on eternally. The paper points out the important role of GLAMS in the era of Digital Intelligence, including the preservation, dissemination and utilisation of cultural heritage resources, as well as the exploration of ultimate human problems.

The Information Processing System for Chinese Characters (also known as “Project 748”) is a national key scientific and technological research project initiated by China. “Project 748” aims to change the backward situation of China’s print industry and solve the problem of computer information processing for Chinese Characters. “Project 748” has promoted the all-round development of information processing technology for Chinese Characters, and also promoted the informatization progresses of library and information cause, which is a milestone in the development history of library and information technology in China. With historical research method and network survey method, this paper systematically reviews the development history of “Project 748” and one sub project “Intelligence Search Using Chinese Characters”, analyzes its important significance in the history of library and information technology in China, and provides references for the research and development of library and information technology in China.

Accurately grasping the current application status of AI technology in university libraries has important reference value for promoting service innovation and improving the level of intelligent services in university libraries. Using a combination of literature research, online research, and questionnaire surveys, this study aims to investigate the current application status of AI technology in university libraries and technology vendors nationwide. The results of the survey are analyzed for both achievements and problems regarding the research topic. The paper clarifies current situation of the development prospects, practical status, and implementation obstacles of AI technology in university libraries, and proposes practical suggestions to promote the development of AI technology in university libraries from three aspects: internal and external cooperation, library leaders, and librarians.

In the context of rapid information technological advancement, the museum industry is undergoing profound transformation. The integration of digital technologies, particularly artificial intelligence, has not only reshaped the presentation of exhibits but also revolutionized visitor interaction experiences. Placing visitors at the center has always been a cornerstone of museum service philosophy. The development of smart guided tour systems, as a best practice of the convergence of technology and humanistic values, is leading new directions in museum education, communication,and service experiences. This paper clarifies the current state of museum guided services, the design of guided functions, and their application value. It further analyzes the challenges and proposes optimization strategies for the construction of smart guided tour systems in museums, aiming to provide insights and strategies for the sustainable development of museums.

With the rapid advancement and widespread application of artificial intelligence (AI) technology, libraries-an essential institution for knowledge preservation and public service-are encountering unprecedented opportunities and profound challenges. AI holds the potential to significantly enhance operational efficiency and deliver more personalized,intelligent services to users. However, issues such as industry transformation, practical difficulties in technology implementation, and the adaptation of librarians’ skills have also emerged as pressing concerns for the library sector.This paper draws on the core discussions from the Directors’ Forum at the 21st Advanced Digital Library Seminar (ADLS 2025) and integrates insights from the industry and the academic to systematically explore new development pathways and strategies for libraries in the AI era, covering dimensions such as resource development, service models, spatial design, and professional competencies. The study aims to provide theoretical references and practical guidance for libraries undergoing transformation in this new era, ensuring their sustainable development and value redefinition amid the AI wave while maintaining their critical role in an intelligent information environment.

As one of the core parts of the ancient Chinese political system, the system of prime ministers is an important tool for exploring the top-level institutional design of the ancient China and deepening the understanding of the operation of the ancient regime. At present, there are a lot of researches related to prime ministers, but they are mostly focused on the whole system, with less consideration of geography as an important influence factor, and lack of trajectory portrayal of the changes in the positions of the prime ministers since they joined the civil service. This paper takes the prime ministers of the Song Dynasty as an example, and based on the database of Chinese biographies and historical documents, systematically analyzes the career changes and geographic distribution of prime ministers, and constructs a map system of the trajectories of Chinese prime ministers. Through the complete interactive map of prime ministers’ career trajectories and the website search function, the research provides an intuitive research and learning tool for historians, geographers and the public, which helps to deepen the understanding of the political geography of ancient China, and provides methodological support for the in-depth application of the Historical Geographic Information System (HGIS) to the study of humanities fields.

With the rapid development of artificial intelligence and large model technologies, the global library industry is gradually integrating these technologies into business processes and user services. As a key technical architecture, AI agents not only address the shortcomings of large models in library business understanding and tool cognition but also provide an important technical path for the intelligent development of library industry goals and tasks through the combination of domain expertise and data. This article showcases the value of AI agent technology in optimizing business processes and enhancing user experience by investigating the intelligent practices of libraries at home and abroad, as well as the application of AI agents in other industries such as publishing, museums, and education. The article first analyzes the recent trends in AI agent technology development and the problems encountered when existing large model technologies are applied in real-world scenarios. Then, it discusses how the existing AI agent application experiences can provide insights for library innovation and summarizes the paths for applying AI agents in optimizing business processes. Finally, the article proposes a framework for optimizing library business processes based on the AI agent architecture and introduces a preliminaryexploration of applying AI agents to a specific scenario, “Duty Assistant.” It analyzes the potential application of AI agents in libraries in the future. The research in this article indicates that AI agent technology can drive the intelligent transformation of libraries, achieve seamless integration of new te。chnologies with traditional systems, and lay a solid foundation for future knowledge services.

The new wave of artificial intelligence (AI) technology is profoundly propelling human society towards comprehensive transformation, and a new era of AI-driven scientific research (AI for Science) is on the horizon. The paradigm of scientific research is undergoing an unprecedented acceleration in transition. This article analyzes the concept and research significance of AI for Science, reviews the development status of AI for Science from basic disciplines such as life sciences, materials sciences, physics, mathematics, chemistry, and earth sciences, discusses the transformative role of artificial intelligence in scientific research, and explores its guiding significance for future research directions. It also discusses the challenges faced in areas such as scientific research data, ethical safety, computational resources, and talent cultivation, in order to provide strategic insights and response strategies for the future development of the scientific research field.

With the rapid development of general artificial intelligence large model technology, humanoid robots, as a cutting-edge achievement integrating artificial intelligence and physical robotics, have attracted widespread attention.Focusing on the theme of library humanoid robots, this paper systematically reviews and analyzes their current development status and existing bottlenecks through methods including web-based survey, literature analysis, and content analysis, from multiple dimensions including academic literature, major products, and application practices. Meanwhile, by tracking the latest technological advancements and development trends, it proposes outlooks on future innovation paths and implementation strategies for library humanoid robots, aiming to provide references for the theoretical and practical development of smart libraries.

The evolution of library information retrieval has progressed from manual card retrieval and electronic catalog retrieval to full-text retrieval. Simultaneously, the forms of library collections have undergone significant transformations,transitioning from physical collections to various databases and, ultimately, to a digital resource system encompassing diverse modalities. In the era of large language models, library information retrieval is advancing towards intelligence. Large language models can handle natural language queries from complex data sources, enhancing the retrieval methods. Based on a review of the historical development of retrieval technologies, this paper explores the enhancements brought by large language models to library information retrieval, proposes a library information retrieval system architecture based on Retrieval-Augmented Generation (RAG). It also analyzes the impact of large language models on future library information retrieval patterns.

The revolution in information technology has driven profound transformations in the library industry and its information systems, with large language models exerting a disruptive impact on knowledge production and services in libraries. This paper analyzes how large models empower libraries in knowledge production, knowledge services, and knowledge experiences, and proposes that the library industry should restructure next-generation library service platforms. Based on industry research and practice, the library should formulate a roadmap for applying large models, actively promote their deep integration into technical architectures, strengthen data governance capabilities, enhance API automation platform development, and shift the overall pattern of platform architecture. At the same time, collaborative innovation within the library community will also be the key driving force behind this transformation.