Journal of Library and Information Science in Agriculture >
Empowered Digital Reading Promotion of Historical Documents with Generative AI
[Purpose/Significance] With the growing demand for intelligent cultural services, libraries are seeking innovative approaches to enhance access to and engagement with historical literature. Generative AI presents promising opportunities for transforming digital reading services, particularly in processing, interpreting, and promoting complex historical documents. This study investigates the integration of generative AI into library-based historical literature promotion, aiming to address persistent access limitations, foster more interactive user experiences, and optimize the depth and breadth of reading engagement. [Method/Process] This research adopts a multi-method approach combining literature review, comparative platform observation, and empirical implementation practice. The study focuses primarily on Shanghai Library's historical digital collections and AI-enabled services. It develops a structured three-layered implementation framework encompassing the data layer, application layer, and service layer-each mapped to corresponding technical and operational phases of digital reading promotion. Within this architecture, a six-step service pathway is articulated: demand analysis, activity planning, content mining, multimodal interaction, content review, and intelligent recommendation. Extensive practical experimentation is conducted across these stages. Key innovations include the application of Retrieval-Augmented Generation (RAG) to support complex historical document Q&A; the use of multimodal creative tools (e.g., Midjourney) to generate engaging visual materials; implementation of voice-based AI interactions to improve accessibility for diverse user groups; and the deployment of dynamic content management modules for librarians to curate and monitor AI-generated materials. Additionally, backend tools such as user profiling dashboards, personalized push notification systems, and topic-based knowledge repositories are developed and tested to enhance librarians' ability to deliver targeted and data-driven reading promotions. [Results/Conclusions] The findings demonstrate that generative AI significantly enhances the efficiency, precision, and user engagement levels of historical literature services. AI-driven methods substantially improve OCR accuracy, streamline metadata generation, facilitate both visual and semantic content creation, and enable real-time interactive services via natural language interfaces. These advancements contribute to a more immersive and responsive digital reading experience. However, several challenges persist, including limited availability of domain-specific training data, the ongoing risk of AI-generated content inaccuracies (hallucinations), and unresolved intellectual property considerations. The study emphasizes the importance of developing domain-specific large language models, establishing expert-assisted validation mechanisms, and formulating clear legal and ethical guidelines for AI-generated content in the library context. While the prototype platform developed in this research exhibits notable gains in user engagement and librarian workflow support, its long-term sustainability hinges on fostering cross-institutional resource collaboration, advancing supportive policy frameworks, and embedding robust ethical safeguards. Future research directions include the exploration of adaptive AI training systems incorporating user feedback loops, integration of cross-library data resources, and the enhancement of multilingual AI capabilities to better serve diverse and global user communities.
TAN Miao , DAI Mengfei . Empowered Digital Reading Promotion of Historical Documents with Generative AI[J]. Journal of Library and Information Science in Agriculture, 2025 , 37(4) : 83 -93 . DOI: 10.13998/j.cnki.issn1002-1248.25-0217
1 |
陈立. 历史文献的阅读推广与可持续发展[J]. 国家图书馆学刊, 2015, 24(1): 46-51.
|
2 |
傅宝珍. 知识服务背景下古籍VR阅读推广研究[J]. 图书馆工作与研究, 2022(1): 108-115.
|
3 |
杨俊, 谭丰隆, 陈婧. 从ChatGPT到“LibGPT”: 生成式人工智能驱动的新一代图书馆[J]. 图书情报工作, 2024, 68(9): 3-12.
|
4 |
赵瑞雪,黄永文,马玮璐,等.ChatGPT对图书馆智能知识服务的启示与思考[J].农业图书情报学报,2023,35(1):29-38.
|
5 |
安子栋, 敬卿, 郝志超, 等. 基于生成式AI技术的图书馆文献资源管理创新策略[J]. 图书馆工作与研究, 2023(S1): 9-16.
|
6 |
黄爱平. 历史文献学学科基础理论与教材编写的思考[J]. 文献, 2013(1): 3-10.
|
7 |
茆意宏. 数字阅读推广的概念、机制与模式[J]. 图书情报知识, 2020, 37(2): 51-59.
|
8 |
周笑盈. 虚拟现实技术在古籍智慧化阅读推广中的应用与实践[J]. 农业图书情报学报, 2022, 34(8): 79-91.
|
9 |
|
10 |
|
11 |
|
12 |
International Federation of Library Associations and Institutions. Using innovative technologies to reimagine libraries and archives services in the National Library Board Singapore[EB/OL]. [2025-03-16].
|
13 |
北京大学数字人文研究中心. 吾与点古籍智能处理平台[EB/OL]. [2025-03-16].
|
14 |
北京文点益度科技有限公司. 吾与点智能数据平台-智能数据处理解决方案 | 吾与点智能数据平台[EB/OL]. [2025-03-16].
|
15 |
全国报刊索引平台[EB/OL]. [2025-03-16].
|
16 |
周笑盈. 国家图书馆“《永乐大典》VR全景文化典籍”实践探索: 虚拟现实赋能图书馆沉浸式阅读推广的创新路径[J]. 国家图书馆学刊, 2022, 31(6): 80-89.
|
17 |
吴若航, 茆意宏. 生成式人工智能变革图书馆阅读推广研究[J]. 图书与情报, 2023(6): 62-69.
|
18 |
李鹏, 宋西贵. AIGC技术赋能图书馆阅读推广工作的创新应用[J]. 农业图书情报学报, 2023, 35(12): 84-93.
|
19 |
刘琼, 刘桂锋, 王鹏. AIGC赋能图书馆阅读推广智慧服务的框架和应用研究[J]. 图书馆学研究, 2024(2): 108-118, 107.
|
20 |
肖鹏, 邓默言, 苏洁, 朱海缘. 阅读推广标准化建设研究报告(2023年版)[J]. 农业图书情报学报, 2023, 35(10): 34-47.
|
21 |
皇甫娟. AI赋能的图书馆数字阅读推广场景化服务模式研究[J]. 图书馆界, 2022(5): 6-10.
|
22 |
刘琼, 周云峰, 苏文成, 等. AIGC环境下阅读推广规范化管理体系研究[J]. 农业图书情报学报, 2023, 35(10): 48-57.
|
23 |
陈柳红. 人工智能阅读器对儿童阅读效果的实证研究: 以比巴为例[J]. 山东图书馆学刊, 2021(4): 68-71.
|
24 |
宫平. 人工智能在图书馆绘本阅读领域的应用模式探索[J]. 图书馆学研究, 2020(2): 88-92, 101.
|
25 |
王军, 刘成林, 金连文, 等. 系列笔谈之四: 智能时代古籍OCR技术[J]. 数字人文, 2022(3): 95-125.
|
26 |
|
27 |
|
/
〈 |
|
〉 |