This paper elaborates on the strategic vision of the Shanghai Library as it embarks on the development of a new generation research public library at the beginning of the “15th Five Year Plan” period. It reviews the historical evolution of the Shanghai Library since its establishment, when it was positioned as a research public library, and analyzes the strategic background behind the subdivision of its vision during the “14th Five Year Plan” into building both a “world class urban public library system” and a “world class research public library”. By benchmarking against the public library systems of international metropolises such as New York, London, Paris, and Tokyo, the paper proposes forming four major service systems: “serving the general public”“serving professionals” and “serving decision making”, based on six major library facilities, including the Shanghai Library East Branch, the Huaihai Road Branch, and the Xujiahui Library(also known as the Bibliotheca Zi Ka Wei). Against the backdrop of digital and intelligent transformation, the paper argues that the new generation research public library should adhere to the principles of “five equal emphases” and “three reconstructions,” balancing the dual missions of being both a “public library” and a “research library,” and ultimately achieving a leapfrog development toward an intelligent and hybrid library suited to the era of mass reading.

How research public libraries can achieve innovative transformation through the development of thematic libraries is a critical issue in the era of digital intelligence. The Chinese Cultural Celebrities Manuscript Library of Shanghai Library serves as a typical case of such transformation. This paper traces the library's development over three decades(1990-2026), dividing it into three stages and analyzing the transformational characteristics of each stage from four dimensions: collection development, service model, space utilization, and management mechanism. Based on this analysis, four key elements are identified: systematic resource development strategy, research service integrated function repositioning, public cultural space construction for diverse user groups, and data driven operational mechanism. These elements constitute a transformational framework that shifts the focus from “collection” to “utilization”, and extends services from specialized academic support to broader public engagement. The paper further proposes the construction of a research service integrated smart service platform to address the needs of resource datafication, business platformization, and smart application development.

Genealogy service, Shanghai Library, Digital humanities, Public cultural service, Research public library

Taking the Bibliotheca Zi-Ka-Wei of the Shanghai Library as the research subject, this paper investigates the transformation practices of a traditional special collections department into a modern research library. Employing content analysis, case study, and historical textual research, the study traces the historical evolution and developmental trajectory of the Bibliotheca Zi-Ka-Wei, examines its collection resources, public service models, and scholarly achievements, while incorporates an analysis of comparable research libraries abroad. The paper proposes that the transformation of the Bibliotheca Zi-Ka-Wei into a modern research library should be advanced through four key dimensions: collection development, physical space reengineering, service model innovation, and the enhancement of talent cultivation and research capabilities.

AbstractResearch on the information cocoon has generated extensive academic debates over its objective existence, causes, and effects due to the metaphorical nature of the concept and the complexity of its formation mechanisms, resulting in fragmented research findings, and highlighting the urgent need for systematic theoretical integration and clear methodological guidance. To overcome this bottleneck, this paper introduces Marx's “retrospective reasoning” as the core methodology, advocating an approach that begins with the observable effects of information cocoons and works backward to infer their formation logic. Based on this, the study constructs a multi-dimensional collaborative analysis framework integrating “semantic space” and “topological space” to systematically deconstruct the conceptual connotation of information cocoons and guide empirical quantification and resolution research. The findings reveal that, from the “retrospective reasoning” perspective, information cocoons can be redefined as an information-cognitive coupling system encompassing semantic homogenization, structural agglomeration, and dynamic stability. Their quantification issues can be transformed into semantic distance calculations in semantic space and structural feature analysis in topological space. Strategies for mitigating the information cocoon may focus on the destabilization of dynamics topological structures and the reconnection strategies of heterogeneous information bridging. The innovation of this study lies in transforming the information cocoon concept from a metaphor into an operational analysis model through the method of “retrospective reasoning”, thereby providing a clear theoretical framework and methodological orientation for future research in this field.

As an important cultural carrier of cultural heritage, special collection spaces in libraries face multiple challenges in data governance, such as multimodal integration, dynamic preservation, and service innovation. This article proposes a virtual-real fusion data governance model for library special collection services, centered on digital twin technology and a data governance theoretical framework. This model follows four principles: full lifecycle management, user demand orientation, security and compliance, and scalability and flexibility. At the same time, it constructs a governance framework consisting of six layers: the physical perception layer, data relationship layer, data transformation layer, digital twin core layer, service application layer, and security governance layer. This article takes the Li Zhengdao Library of Shanghai Jiao Tong University as an example to demonstrate how to use digital twin technology to achieve dynamic preservation, knowledge association, and interactive visualization services of its special collections through digital governance and application practices. The research results indicate that digital twin technology can effectively improve the quality and efficiency of library special collection services, enhance user participation, and provide new ideas and methods for the digital preservation and inheritance of cultural heritage.

Microblog serves as a vital platform for public library to provide public services and enhance interlibrary communication. Applying Social Network Analysis(SNA) to explore the structural characteristics of the microblog community and the influence of library accounts can offer valuable insights for optimizing social media operation strategies in public libraries. This study focuses on provincial public libraries to construct a Weibo-based social community network, identifying its structure from three dimensions: network integrity, centrality and subgroups. Centrality metrics were extracted. Combined with basic microblog indicators such as the number of followers and posts, factor analysis was employed to establish an evaluation index system for public library's microblog accounts. The results indicate that the provincial public library microblog community exhibits high connectivity with weak evidence of fragmented subgroups. There is significant variation in the influence of microblog accounts, with the majority demonstrating low influence. 

The library in special education schools serves as both a center for information resources and an essential venue for educational activities. This study investigates the libraries of 24 special education schools at the compulsory education level in Shanghai, and systematically examines the current state of development across various dimensions, including environmental construction, resource allocation, management and operation, and service support. Furthermore, the study addresses how libraries can contribute to the high-quality development of special education and enhance the reading abilities of children with disabilities. Based on the findings, recommendations are made to strengthen the establishment of long-term operational mechanisms, optimize the resource system, innovate service content and formats, deepen integration with educational practices, and enhance professional training for librarians. These suggestions aim to provide insights for improving the educational effectiveness of libraries in special education schools in the digital age. 

In view of the multi-stakeholder nature of open government data, this study introduces the theory of value co-creation to analyze practical cases, and proposes feasible open government data reuse models and mechanisms to help the development of digital economy. From the innovative theoretical perspective, this paper takes Shanghai Public Data Open Platform as the research object, and analyzes and excavates five aspects including open government data reuse, participants, co-creation mode, key factors and core value types. Three models of open government data reuse are summarized: government-enterprise cooperation, open data competition and regional application. It further puts forward an open government data reuse mechanism with the government leading together with academic institutions, commercial institutions, non-governmental organizations, individuals, media and other stakeholders jointly participate in exploring and creating data value. It is suggested that the future work should focus on aspects such as cultivating data reuse awareness, meeting user demands, encouraging multi-stakeholder participation and strengthening practical exploration. 

The rapid advancement of artificial intelligence has increased the complexity of data governance, making multimodal data governance a key approach to addressing these challenges, and calling for a clear articulation of its core issues and governance pathways. Building on a conceptual analysis of multimodal data governance, this study focuses on the governance challenges emerging from the interaction between multimodal data characteristics and AI environments. From the dual perspectives of technological innovation needs and cybersecurity incidents, the study systematically analyzes the internal mechanisms and external manifestations of these issues in terms of their causes and consequences. Accordingly, this study proposes a multimodal data governance pathway in  AI environments that takes value co-creation as the core objective, technological regulation as the key instrument, and a coordinated soft-and-hard regulatory framework as the central model. The findings offer a reference for improving data governance frameworks in AI environments. 

With the acceleration of social development and aging process, the spiritual and cultural needs of elderly populations have become increasingly prominent, making reading an essential component. Understanding the reading needs of elderly readers is helpful for optimizing public cultural services and promoting the in-depth development of reading among the elderly. The borrowing data of elderly readers from libraries in various districts of Shanghai in 2019 was used as a reference. The BERTopic model was employed to conduct topic clustering, and the reading needs of elderly readers were summarized based on the ERG theory. Statistical analysis was carried out on the reading needs of elderly readers across different genders and age groups. The study revealed that elderly readers exhibit diverse reading interests, contrary to traditional stereotypes, covering health and wellness, literature, history, and other themes. Gender significantly influences reading preferences: male seniors tend to read books that satisfy growth-related needs, while female seniors prioritize relationship-focused content. Age shows minimal impact, with relationship-themed reading consistently dominating across all age groups, whereas survival-oriented reading remains relatively low across three age brackets. The research shows that the promotion of reading for the silver age should adhere to a positive view of aging and pay more attention to the life value of elderly readers.

Promoting the sharing of digital dividends among all citizens is a crucial component of advancing common prosperity and achieving the strategic goal of Chinese-style modernization. The study aims to deepen the academic understanding of digital literacy and digital divide by exploring the interrelationship between reading behaviors and digital literacy, to promote the construction of a digital society, and thus to enhance China's international competitiveness in the digital age. The study utilizes the CGSS2017 data to comprehensively formulate an evaluation index system for citizens' digital literacy, establishes an impact model of reading behavior on digital literacy for regression analysis, and further investigates the effects of reading behavior on digital literacy from the perspective of the digital divide. The results show that both dimensions of reading behaviors—reading frequency and reading volume—can positively enhance citizens' digital literacy. Notably, digital reading behavior exhibits a more significant effect on promoting digital literacy compared to traditional paper-based reading. Reading behaviors positively influence citizens' digital literacy through three pathways: self-efficacy effects, economic status effects, and learning effects. Furthermore, reading behavior has a greater promoting effect on the digital literacy of most digitally disadvantaged groups, thus helping to bridge the digital divide to a certain extent. Accordingly, the study suggests that relevant authorities should reassess the value of reading from the perspective of improving digital literacy, fully tap into the potential of reading from the perspective of bridging the digital divide, and carefully regulate publications from the perspective of maintaining digital security. These findings systematically reveal the role and multidimensional mechanisms of reading behavior in enhancing digital literacy. The study explores potential pathways for bridging the digital divide, contributing to the expansion of the digital divide theory, formulating digital literacy improvement strategies, and promoting balanced development in the digital society. 

National reading is a crucial component of building a culturally strong country, and the core role of reading promotion policies has become increasingly prominent. Exploring the operational mechanism of China's reading promotion policies can provide valuable insights for optimizing the formulation and implementation processes of future policies. Based on grounded theory, this paper systematically analyzes 74 national and local reading promotion policy texts, and constructs a theoretical framework for the operational mechanism of China's reading promotion policies, which includes 4 main categories and 12 initial categories. The finding reveals that the policy-driven effectiveness is first released through systematic policy design, laying the foundation for reading promotion actions. Then, an implementation pathway integrating scenario-based practical actions and service guarantee mechanisms is formed to enable policy transformation into practice. Meanwhile, the outcome feedback mechanism based on supervision and evaluation scientifically tests the effectiveness of policy formulation and implementation. Policy driving forces, implementation transformation, and outcome feedback are connected with each other, coordinate dynamically, and jointly promote national reading to transform from an abstract national policy into a public cultural practice that covers all regions and all groups of people. 

To address the needs of knowledge mining tasks, this study conducts a comparative analysis of the performance of large language models (LLMs) for entity and relation extraction, aiming to explore more precise knowledge extraction approaches and provide insights for fine-grained knowledge analysis. Using Chinese and international information science journal articles between 2022 and 2024 as data sources, we design and evaluate a multi-dimensional fusion prompt framework, testing it on mainstream LLMs, including DeepSeek, Kimi Chat Assistant, ChatGPT, and Gemini. Performance is systematically assessed through both strict and relaxed matching evaluation mechanisms. The results demonstrate that the domestic model DeepSeek outperforms others in entity and relation extraction tasks. On Chinese and foreign datasets, its F1 scores surpass competing models by an average of 7.55% and 7.89% in “problem entity” recognition, 8.82% and 15% in “method entity” recognition, and 16.98% and 9.91% in “problem-method” relation extraction, respectively. These findings indicate that DeepSeek exhibits strong capabilities in structured knowledge extraction. The proposed prompt framework and model comparison provide methodological references for LLM-based knowledge discovery research.