10 November 2026, Volume 46 Issue 11
-
Select all|
-
LIBRARY TRIBUNE. 2026, 46(11): 1-20.Directly delivering high-quality cultural resources to the grassroots constitutes a critical strategy to enhance the quality and efficiency of public cultural services and to promote equal and accessible services,with public libraries playing an important role. Focusing on delivering high-quality cultural resources to the grassroots by public libraries,this study uses the grounded analysis method to conduct three-level coding of relevant policy texts,extracting five main categories：value orientation,system support,resource supply,supply methods and supply effectiveness,thereby constructing a policy mechanism model. Using a representative case analysis method,it summarises and analyzes the practical development patterns. The research findings indicate that on the whole,there is high consistency between the policies and practices of delivering high-quality cultural resources directly to the grassroots by public libraries. Basic reading resource supply,service network extension,and digital platform construction have achieved certain results,but there are still deficiencies in the continuous operation of the general-branch library system,the supply of educational and learning resources,the transformation of cultural inheritance resources,and quality-oriented evaluation. Based on this,a policy-practice coupling explanatory model is used to illustrate how public libraries can channel high-quality cultural resources directly to grassroots communities. This reveals the dynamic mechanism of interaction between policy guidance,practical transformation,and feedback optimization,and puts forward corresponding suggestions including refining implementation mechanisms,strengthening the support of the general-branch library system,optimizing the structure of resource supply and service scenarios,and improving the evaluation and feedback mechanism.
-
LIBRARY TRIBUNE. 2026, 46(11): 21-37.Analyzing policy tool selection practices and exploring the effective pathways of policy tool supply are a practical necessity for innovation in public cultural service development. Through content analysis,this study quantifies the characteristics of policy tool selection in 914 public cultural service policies issued by China’s 31 provincial-level administrative units from 2006 to 2024,and then applies fuzzy-set qualitative comparative analysis （fsQCA） to identify the configurational paths through which policy tool supply promotes the development of public cultural services. The study reveals that the supply of policy tools at the provincial level exhibits the characteristics of shared content,differentiated types,structural imbalance,and complementarity and substitutability. The advancement of public cultural services depends on the synergistic configuration effects of mandatory-type,incentive-type,capacity-building-type,persuasive-type,and transformative-type policy tools. The differentiated effects of policy tool configurations give rise to four configurational paths：mandatory-regulation,diversified-incentive,capacity-reinforcement,and governance-coordination. Accelerating the modern transformation and upgrading of the public cultural service system does not necessarily rely on the quantitative accumulation or proportional allocation of policy tools. Rather,decision-making should be grounded primarily in the functional characteristics of policy tools,regional resource endowments,and developmental needs,with emphasis on enhancing the applicability,compatibility,and contextual relevance of policy-tool combinations and their application methods,so as to design precise and adaptive paths that address shortcomings in policy effectiveness.
-
LIBRARY TRIBUNE. 2026, 46(11): 38-54.In response to the imperatives of the market-oriented allocation reform for data elements and the goal of promoting high-quality development of the digital economy through their efficient and orderly circulation,this paper examines the current compatibility between data element circulation forms and policy tools in China from both quantitative and qualitative （content） perspectives. Under a two-dimensional analysis framework of "policy tools- circulation forms",the study coded 261 policies related to data element circulation,and identified the distributional characteristics of three types of policy tools （supply-oriented,demand-oriented,and environment-oriented） and three circulation forms （data openness,trading,and sharing）. Cross-analysis further revealed a quantitative "allocation imbalance" between the two dimensions. Subsequently,it extracted the core demand nodes associated with each circulation form from the literature and interview materials,measured their co-occurrence frequency with policy tools at the paragraph level,and analyzed the "supply-demand mismatch" at the content level. The paper proposes policy optimizations,including the coordinated application of environment-oriented tools,the increase of supply-oriented tools,and the addressing of weaknesses in demand-oriented tools,with the aim of transforming data element circulation policy tools from an extensive model of "generalized coverage" to an orderly regulatory framework grounded in "precise adaptation".
-
LIBRARY TRIBUNE. 2026, 46(11): 55-65.The rise and fall of Peking University Founder Group’s Apabi,an iconic platform in China’s digital publishing and digital library sector,offers a representative example for examining the evolution of the industry. Through literature review and interviews with practitioners,this paper systematically traces the three-stage development trajectory of Apabi：initial exploration,innovation and maturity,and eventual obsolescence. It analyzes the whole-process digital publishing solution Apabi built around DRM copyright protection and CEB/CEBX fixed-layout technologies,and summarizes its early experiences in institutional services,copyright governance,and industrial chain integration. Reflecting on the technology-market synergy,independent operation,industrial-chain responsiveness,business layout,and copyright balance,the study pinpoints the key factors behind Apabi’s decline from prosperity：namely its failure to translate technological advantages into market competitiveness,over-reliance on its parent company,a narrow business portfolio,excessively strict copyright controls,and a delayed response to industrial changes. Apabi’s history demonstrates that sustainable development in digital publishing can only be achieved by maintaining a dynamic balance among technological innovation,market adaptation,and social demand,and reconciling copyright protection with universal access to knowledge. Its experience and lessons provide valuable insights for today’s e-book and digital publishing industries.
-
LIBRARY TRIBUNE. 2026, 46(11): 66-77.As a new data infrastructure that enhances data interoperability,governance capabilities,and data value co-creation,trusted data spaces offer libraries a crucial opportunity to overcome practical challenges like data silos and slow value growth and to reshape their service ecosystems. Drawing on symbiosis theory,this article conceptualizes the library trusted data space as a symbiotic system of data providers,consumers,service providers,and regulators,covering libraries,publishers,and end-users. From the perspectives of symbiotic units,symbiotic models,and symbiotic environments,the study examines the underlying mechanisms of stakeholder collaboration,value co-creation,and ecological evolution over the course of the space’s development. Based on the theory of value co-creation,it proposes a progressive implementation pathway that focuses on interaction momentum,value co-creation,and innovative vitality,with the aim of providing practical methodological guidance for building trusted data spaces in libraries.
-
LIBRARY TRIBUNE. 2026, 46(11): 78-91.The rapid development of generative artificial intelligence poses structural challenges to the current copyright legal system. Digital humanities （DH） platforms simultaneously serve as aggregators of digitized cultural resources,data infrastructure providers for text and data mining （TDM） and AI training,and integrators of AI-enhanced research tools. This triple role places them at the critical intersection where humanities research and copyright protection converge. Clarifying the challenges they face in copyright governance and proposing solutions is of great significance for safeguarding academic freedom and ensuring open access to cultural heritage. Through comparative analysis and case study,this paper examines legislation,judicial precedents,and policy documents across three major jurisdictions — China,the United States,and the European Union. Using an "input-process-output" analytical framework,it systematically analyzes the threefold dilemma confronting DH platforms：at the input end,the unclear copyright status of digitized public domain cultural heritage works and the "contractual veto" imposed by publisher licensing agreements;at the process end,the blurred legal boundaries of TDM and the ownership vacuum regarding intermediate outputs;and at the output end,the copyright risks associated with generated content and the escalating risks of platform liability. The paper proposes a three-pronged approach — legal and policy advocacy,institutional development,and international collaboration — arguing that DH platforms must shift from passive enforcers of copyright compliance to proactive builders of DH intellectual property rules. In doing so,they can secure an irreplaceable governance role within the emerging intellectual property order of the AI era.
-
LIBRARY TRIBUNE. 2026, 46(11): 92-101.The rapid development of large language models （LLMs） aligns well with the demand for intelligent transformation of academic services in university libraries,while also providing new technical support for libraries to enhance the efficiency of knowledge services and expand intelligent service scenarios. Using DeepSeek as a case study,this paper applies grounded theory to qualitatively analyze interview data and publicly available texts from 33 university libraries,revealing the core categories and application logic of LLM-empowered academic services in university libraries. The study finds that DeepSeek empowers academic services in university libraries through five key dimensions：core academic service capabilities,adaptation to diverse service scenarios,innovation in implementation models,building a service operation ecosystem,and ethical and security governance of academic services,thereby forming an application logic model of "technology empowerment - scenario adaptation - value realization - risk governance". Based on the results of grounded theory analysis and the actual application scenarios of university libraries,the study proposes optimization strategies from the perspectives of technical infrastructure,service systems,implementation approaches,and governance frameworks,with the aim of providing insights for the intelligent transformation of academic services in university libraries within the context of domestically developed LLM adoption.
-
LIBRARY TRIBUNE. 2026, 46(11): 102-113.Virtual community digital collectibles represent a new form and scenario for the development,utilization,and consumption of cultural resources in the context of cultural digitization. This study employs in-depth interviews and the grounded theory method to explore the factors influencing purchasing decisions among users of digital collectibles in virtual communities,as well as the mechanisms underlying these decisions,with the aim of promoting the healthy development of digital collectibles services and consumption in the context of cultural digitization. The study found that the primary factors influencing users’ purchasing decisions regarding digital collectibles in virtual communities fall into three major categories：cultural drivers,technological experience,and economic perception. Specifically,these include cultural emotional resonance,technological trust,value assessment and so on. Among these,cultural drivers serve as the internal driving force behind purchasing decisions for digital collectibles in virtual communities,directly influencing the formation of purchase intent;while technological experience and economic perception serve as moderating factors in the transition from purchase intention to actual purchase behavior,potentially influencing the probability and intensity of the relationship between purchase intention and purchase behavior. The "Cultural-Economic-Technological Integration Model" proposed in this paper reveals the complex mechanisms underlying user consumption of digital collectibles,providing theoretical foundations and practical insights for the design of cultural empowerment in digital collectibles,platform functionality optimization,and information service strategies.
-
LIBRARY TRIBUNE. 2026, 46(11): 114-126.Generative artificial intelligence has made conversational search a novel paradigm for health information retrieval,yet systematic comparative studies with traditional search paradigms remain insufficient. This study employs a mixed-methods design that integrates behavioral experiments,questionnaire surveys,and semi-structured interviews to explore the behavioral differences and learning effects between the two search modes in health information seeking. A total of 60 users were randomly assigned to either a conversational search group （Doubao） or a traditional search group （Baidu）. Search experiments were conducted across three health-related scenarios involving six tasks,and data were collected through screen-recording logs,concept map evaluations,and questionnaires. The results reveal that conversational search significantly outperforms traditional search in information integration efficiency and response quality,and delivers better objective learning effects,with its advantage growing with task complexity. In contrast,traditional search has an edge in users’ perceived credibility of information sources. The two search modes are functionally complementary,suggesting that their use should be tailored to task complexity.
-
LIBRARY TRIBUNE. 2026, 46(11): 127-139.Traditional multimodal misinformation detection struggles to provide a clear chain of evidence and an interpretable decision-making process. In light of this,this paper aims to explore a multimodal misinformation detection method that combines detection performance,interpretability and practical feasibility. Addressing four typical forms of multimodal misinformation—namely,“both text and images are fake” “single-modal image forgery” “single-modal text forgery” and “text-image mismatch”—this paper proposes a dual-path,evidence-driven multimodal misinformation detection framework based on large language models. Leveraging the powerful comprehension,reasoning and generation capabilities of large language models,this framework constructs a comprehensive verification process across three levels：text authenticity assessment,image authenticity assessment and “text-image” consistency matching. It achieves the authentication of multimodal information through the acquisition of “text-image” dual-path evidence and joint reasoning. Performance metrics and case study results demonstrate that the proposed framework can effectively identify multimodal misinformation in real-world social media scenarios,whilst providing clear and logically sound explanations for its detection. This research offers a real-time,practical method for detecting multimodal misinformation in public opinion monitoring,whilst also validating the application potential of large language models in multimodal information detection.
-
Multi-Embedding Model Fusion-Driven Topic Evolution and Feature Mining：Advances in AI Agent ResearchLIBRARY TRIBUNE. 2026, 46(11): 140-160.Focusing on AI Agent-related papers from four databases,namely arXiv,Scopus,Web of Science （WoS）,and China National Knowledge Infrastructure （CNKI）,this paper attempts to address the imbalanced corpus sizes across databases in a multilingual setting and the difficulty of objectively evaluating the performance of individual embedding models. It proposes an adaptive analytical framework for dynamic topic evolution,specifically designed for multi-source heterogeneous literature. Following the preprocessing of the literature data,the framework employs various embedding models and adaptive algorithmic parameters to model the dynamic topic evolution of the research data. Subsequently,a large language model is adopted to refine and align the extracted topic labels. Through cross-source topic similarity and the first appearance time of topics,the framework delineates the structural differences,evolutionary trajectories,and temporal correlations of AI Agent research across the four databases. The results show that different embedding models affect topic modeling outcomes,and AI Agent research exhibits distinct thematic structures across databases. Specifically,arXiv prioritizes the exploration of frontier technologies,Scopus tends toward engineering applications,WoS emphasizes theoretical development and integrative research,and CNKI highlights application scenarios such as education and knowledge services. There is a temporal stratification across the databases：research foundations for traditional agents,reinforcement learning,and multi-agent coordination were established in the pre-LLM era,and these topics have persisted in subsequent studies. arXiv incorporated LLM-based approaches into AI Agent research earlier than the others,while Scopus,WoS,and CNKI subsequently extended research into engineering applications,theoretical development,and governance regulations. Overall,the evolutionary trajectory of AI Agent research topics follows a progressive and incremental pattern characterized by the parallel continuation of traditional research and the diffusion of LLM-driven approaches.
