17 May 2026, Volume 46 Issue 03

    

• Select all
  |

• Select
  Maintaining Core Principles While Pursuing Innovation: Theoretical Reconstruction and Methodological Renewal in Information Resource Management

  Kang Lele Shi Jing Li Yue Wang Jiajie

  Library & Information. 2026, 46(03): 17-26. DOI:10.11968/tsyqb.1003-6938.2026028

  Abstract ( ) Download PDF ( ) Knowledge map Save

  The construction of the independent knowledge system has emerged as a central priority in Information Resource Management (IRM) discipline. Following the principle of “maintaining core principles while pursuing innovation”, the discipline has achieved a series of advances in both theoretical reconstruction and methodological innovation. This paper adopts integrative re-innovation as its guiding concept to trace the evolutionary process and the internal logic of this endeavor along the above two dimensions. On the theoretical side, we examine three constitutive elements: core concepts, explanatory mechanisms and field boundaries. Among them, core concepts have evolved from carrier-dependent formulations through integrative reconceptualization to data-intelligence-driven reconstruction; explanatory mechanisms have shifted from descriptive and static accounts to dynamic, multi-level frameworks; field boundaries have expanded from carrier-oriented to service-oriented perspectives, and from disciplinary depth to interdisciplinary integration. On the methodological side, endogenous research demands and exogenous technological disruptions, particularly big data and artificial intelligence, have jointly driven paradigmatic change, elevating these technologies from instrumental tools to cognitive enablers and introducing dynamic modeling, complexity analysis, and experimental approaches. We argue that theoretical reconstruction and methodological renewal operate synergistically, together driving knowledge accumulation and paradigmatic evolution in IRM's independent knowledge system in China.
• Select
  Independent Foundation and Open Integration: Reflections on the Construction of the Disciplinary System for the Independent Knowledge System of China’s Information Resource Management Discipline

  Li Yang Zhu Zihan Tao Tianxia

  Library & Information. 2026, 46(03): 27-38. DOI:10.11968/tsyqb.1003-6938.2026029

  Abstract ( ) Download PDF ( ) Knowledge map Save

  The disciplinary system constitutes a core component in the construction of an independent knowledge system for the Information Resource Management (IRM) discipline in China. This article reviews the current state of the IRM disciplinary system in China and argues that its construction and development exhibit a dual character of endogenous evolution and exogenous construction. In response to the new strategy and emerging demands for building an independent knowledge system for philosophy and social sciences in China, the article examines the essential connotations of the disciplinary system. It contends that the IRM disciplinary system that it possesses distinctive features in terms of intellectual foundations, top-level institutional arrangements, in-depth exploration along the information chain, and value construction. It further proposes that innovation in the IRM disciplinary system should be oriented toward maintaining a distinctive focus on information resources, promoting deeper integration between technology and the humanities, and strengthening its capacity for interdisciplinary innovation. Looking ahead, the article puts forward several considerations from the perspectives of disciplinary goals, disciplinary structure, and disciplinary identity.
• Select
  Practical Patterns and Path Exploration of Information Resources Management Oriented towards National Major Strategies

  Wang Liuhong Ba Zhichao Liu Zujun

  Library & Information. 2026, 46(03): 39-52. DOI:10.11968/tsyqb.1003-6938.2026030

  Abstract ( ) Download PDF ( ) Knowledge map Save

  In the context of accelerating the establishment of an independent knowledge system for Chinese philosophy and social sciences, Information Resources Management (IRM) is confronting substantial opportunities and challenges in its transformation from traditional information services to underpinning the modernization of national governance. Focusing on core national strategies such as “Digital China”, “sci-tech self-reliance and self-strengthening”, and “National Security Governance”, this article examines the functional orientation and evolving demands of the IRM discipline concerning strategic importance, collaboration, timeliness, and security. It identifies the adaptive challenges faced by IRM theory amidst changing scenarios and distills the practical logic and service model of IRM with distinctive Chinese characteristics. Based on the “dual-chain and one-axis” theoretical framework,it further investigates a strategically embedded pathway driven by the synergy of “system-resource”. It advocates for a paradigm shift in IRM research from “passive resource provision” to “active strategic support” across four dimensions: top-level strategic design, core competency enhancement, service platform development, and service ecosystem nurturing.
• Select
  Constructing an Independent Knowledge System for Digital Humanities in China: Theory, Method, and Practical Innovation

  Shi Junyi Liu Xinxue Li Jialong Song Ningyuan Pei Lei

  Library & Information. 2026, 46(03): 53-65. DOI:10.11968/tsyqb.1003-6938.2026031

  Abstract ( ) Download PDF ( ) Knowledge map Save

  Constructing an independent knowledge system for digital humanities in China is an essential response to national strategic imperatives and a catalyst for academic paradigm transformation. Grounded in an understanding of China’s own independent knowledge system, this paper reviews the current trajectory of digital humanities within the Chinese context. Utilizing social network analysis and systematic literature review, the study reveals the localization process of the field. It further examines conceptual and theoretical innovations centered on "smart data," "intelligent computing," and "digital memory," alongside the evolution of methodological systems in resource processing, knowledge organization, and cultural communication. The findings suggest that, by rooting itself in China's historical and cultural traditions, the field is transitioning from purely technology applications toward a systematic theoretical framework. Guided by the integration of Marxist principles with fine Chinese traditional culture, the formation of this independent knowledge system will facilitate the integrated advancement of theory, methodology, and practice.
• Select
  Research on the Integrated Paradigm of Knowledge Production and Dissemination in the Age of Artificial Intelligence

  Hu Xiao Chen Changfeng

  Library & Information. 2026, 46(03): 66-73. DOI:10.11968/tsyqb.1003-6938.2026032

  Abstract ( ) Download PDF ( ) Knowledge map Save

  Artificial intelligence is reshaping the long-standing structural separation between knowledge production and knowledge dissemination, driving their integration in specific domains. Through a literature review and case analysis of applications such as AlphaFold2, Consensus, BloombergGPT, this paper draws on paradigm theory and actor-network theory to examine the mechanism of change following AI’s intervention in the knowledge ecosystem. The study finds that AI for Science (AI4S) and AI-Generated Content (AIGC) do not share a single technological architecture, yet both rely on data-driven methods, deep learning, and generative modeling to reorganize the traditional division of labor in knowledge generation, processing, verification, and diffusion. Specifically, by compressing intermediary links, blurring the boundaries between production and dissemination, and reshaping the logic of supply-demand connections, AI transitions knowledge operation from temporal separation to spatiotemporal synchronization, from linear division of labor to functional interweaving, and from static products to dynamic services. However, this transformation is constrained by disciplinary modes of validation, institutional norms, and technological reliability, and gives rise to governance issues such as accountability in algorithmic gatekeeping, the reconstruction of knowledge legitimacy, and the amplification of bias.
• Select
  Simple Words, Profound Truths: A Study on the Explainability of AI Doctor Information and Elder Adults' Health Information Adoption Intentions

  Zhang Ning Yue Dapeng Yuan Qinjian

  Library & Information. 2026, 46(03): 74-87. DOI:10.11968/tsyqb.1003-6938.2026033

  Abstract ( ) Download PDF ( ) Knowledge map Save

  AI doctors are becoming key channels for elder adults to access health information and receive medical assistance. However, elder adults often experience cognitive dissonance and information adoption barriers during interaction with intelligent agents. To adress this, this article explores the mechanism by which the explainability of AI doctor information influences elder adults' willingness to adopt health information. Based on the theory of processing fluency, two parallel experiments were conducted. Study 1 employed a between -subjects design to examined how linguistic fluency and explanation style influenced AI doctor information adoption across different age groups. Study 2 further introduced cognitive dissonance and information efficacy as mediators variables to construct a chained mechanism model, analyzing the psychological mechanism of elder adults in the process of health information adoption. Results indicate that compared to younger adults, elder adults exhibit greater sensitivity to AI-generated explanatory information. Both high linguistic fluency and deep, deliberative thinking explanatory styles significantly enhance elder adults' adoption willingness, with fluency exerting a stronger effect. The explainability of AI doctor information influences older adults' adoption willingness through a chained mediating pathway: “reducing cognitive dissonance—enhancing information efficacy.” This study reveals the complete psychological pathway through which AI doctor information explainability influences older adults' health decisions, offering recommendations for age-friendly design and application of medical AI agents.
• Select
  Identification of Healthcare Topics in Online Consultations Based on Large Language Models: Driven by Doctor-Patient Communication Texts

  Zhang Jiantong Yu Wenjie Li Junchang Zhang Yuting Wang Le

  Library & Information. 2026, 46(03): 88-102. DOI:10.11968/tsyqb.1003-6938.2026034

  Abstract ( ) Download PDF ( ) Knowledge map Save

  In real-world doctor-patient communication, patient self-report texts are overly subjective while physician response texts are overly specialized. Regarding the issue of identifying healthcare topics in such text-based communication, this study leveraged the advantages of Large Language Models (LLMs) and designed comprehensive prompts for online consultations topic identification. By embedding a medical topic taxonomy, a professional terminology database, and a keyword repository within a hybrid framework of multiple prompt strategies, the designed prompt was deployed to compare the consistency of topic identification across diverse combinations of five heterogeneous LLMs (GPT-4o-mini, ERNIE 4.0, DeepSeek-V3, ChatGLM-4, and Qwen-Plus). Finally, three integration schemes for heterogeneous LLMs were proposed: majority voting, vote-excluding-one, and full-consensus. A comparative analysis was conducted on the recognition performance of a single LLM and three schemes based on an expert-annotated dataset. The results indicate that the designed prompts demonstrate high effectiveness, generalizability, and superiority. The healthcare topics such as medication advice and surgical judgment exhibit relatively high consistency, whereas the consistency of treatment advice is relatively low.  GPT?4o?mini performs well among single LLMs, and majority voting exhibits favorable recognition performance.
• Select
  Research on the Configuration Path of Cultural Heritage Data Element Value Release

  Zhang Yunzhong Sang Mengyu

  Library & Information. 2026, 46(03): 103-114. DOI:10.11968/tsyqb.1003-6938.2026035

  Abstract ( ) Download PDF ( ) Knowledge map Save

  Investigating the pathways of factors influencing the value release of cultural heritage data elements is of great significance for fully realizing their value and advancing the digital preservation and innovative utilization of cultural heritage. From the perspective of data productivity and production relations, this study first constructs a factor framework influencing the value release of cultural heritage data elements. It then employs the fuzzy-set qualitative comparative analysis (fsQCA) method to analyze 35 cases of value release within the cultural heritage domain, exploring the condition combinations conductive to high-level value release. The study reveals that the application of advanced data labor tools is a necessary condition for effectively unlocking the value of cultural heritage data elements. The six configuration paths leading to high-efficiency value release are further synthesized into four models: data worker incentivization, circulation and reuse appreciation, market-government coordinated allocation, and data consumption-supply matching.
• Select
  Trusted Data Element Circulation for Deep Integration of Scientific and Technological Innovation with Industrial Innovation: Conceptual Definition, Mechanism Construction, and Enablement Pathways

  Gao Zhihao Zheng Rong Li Fengyun Chen Siran Sun Chuanfeng

  Library & Information. 2026, 46(03): 115-129. DOI:10.11968/tsyqb.1003-6938.2026036

  Abstract ( ) Download PDF ( ) Knowledge map Save

  The deep integration of scientific and technological innovation with industrial innovation faces challenges such as information asymmetry between the innovation  and the industrial chains, barriers to cross-subject data circulation, and the reluctance to share high-value data. Trusted data element circulation provides critical support for the precise alignment between technological supply and industrial demand. Focusing on the task objectives of deep integration, this paper constructs a trusted data element circulation mechanism from four dimensions: trusted subjects, trusted data, trusted technology, and trusted processes. Taking the intelligent data annotation industry as a typical scenario, it further proposes a value release pathway consisting of “innovation alliance, trusted products, trusted space and value feedback.” It is found that through subject collaboration, data supply, technological support, and process governance, trusted data element circulation can promote the precise alignment between technological supply and industrial demand, thereby facilitating the industrial application of scientific and technological achievements and forming a value closed loop for integrated innovation.
• Select
  Research on the Generation Logic and Risk Governance of Deepfake Intelligence in the Generative AI Environment

  Zhang Tao Chen Yun

  Library & Information. 2026, 46(03): 130-141. DOI:10.11968/tsyqb.1003-6938.2026037

  Abstract ( ) Download PDF ( ) Knowledge map Save

  With the rapid development of Generative Artificial Intelligence (GAI), the continuous upgrading deepfake technology has given rise to a new type of threat known as "deepfake intelligence." This not only introduces unprecedented risks to intelligence work but also poses severe challenges to social stability and national security. However, existing research lacks systematic governance solutions tailored to the specificities of the intelligence domain. To this end, this paper constructs a "three-dimensional, four-layer" analytical framework to elucidate the generative logic, core characteristics, and operational mechanisms of deepfake intelligence. The study finds that deepfake intelligence follows a generative pathway of "modality collaboration-semantic fusion-context adaptation," with its risks permeating five levels: technology, intelligence processes, organizations, law, and strategy. On this analysis, the paper proposes a progressive governance strategy encompassing technical countermeasures, process re-engineering, institutional strengthening, legal norms, and international collaboration.
• Select
  Aspect-Level Sentiment Analysis for Online Book Reviews

  Wang Yichun Liu Lu Yang Yi Gao Feng

  Library & Information. 2026, 46(03): 142-152. DOI:10.11968/tsyqb.1003-6938.2026038

  Abstract ( ) Download PDF ( ) Knowledge map Save

  Conducting aspect-level sentiment analysis on book reviews facilitates the identification of users' emotional appeals across various evaluation dimensions, thereby providing decision-making intelligence for book publishing, library collection optimization, and personalized recommendation services. This holds practical significance for promoting the intelligent development of library and information services. This paper constructs an aspect-level sentiment analysis model named ELS-A, based on stacking ensemble learning. A domain-specific book review dataset is built, and the ALBERT (A Lite BERT) model is introduced to generate contextual semantic representations. Heterogeneous base learners, including SVM, CNN, and BiLSTM, are integrated to achieve collaborative modeling of multi-level sentiment features. Experimental results on the Book, Review, and Blog datasets demonstrate that the proposed model outperforms single models and traditional combination methods in terms of accuracy and F1-score, validating the effectiveness of the ensemble learning strategy in improving the accuracy and stability of aspect-level sentiment intelligence extraction from book reviews.