| 付雅明,孙炫宇.GenAI赋能古籍数字化的可信度评估与可信治理策略研究[J].数字图书馆论坛,2026,22(5):83~90 |
| GenAI赋能古籍数字化的可信度评估与可信治理策略研究 |
| Research on Trustworthiness Assessment and Credibility Governance Strategy of Ancient Book Digitization Empowered by GenAI |
| 投稿时间：2026-03-27 |
| DOI：10.3772/j.issn.1673–2286.2026.05.008 |
| 中文关键词: GenAI；古籍数字化；可信风险；可信度评估；可信治理 |
| 英文关键词: GenAI; Digitization of Ancient Books; Trustworthiness Risks; Credibility Assessment; Credibility Governance |
| 基金项目:本研究得到教育部人文社会科学研究一般项目“基于生成式人工智能的数字文化遗产叙事模式与可信治理研究”（编号：25YJCZH049）资助。 |
| 作者 | 单位 | | 付雅明 | 上海大学文化遗产与信息管理学院 | | 孙炫宇 | 上海大学文化遗产与信息管理学院 |
|
| 摘要点击次数: 40 |
| 全文下载次数: 46 |
| 中文摘要: |
| 生成式人工智能（Generative Artificial Intelligence，GenAI）在赋能公共图书馆古籍数字化的同时，也带来了数据失真、模型“幻觉”与权责模糊等可信风险，故亟须加强系统性可信治理。本研究首先分析GenAI赋能古籍数字化过程中引发的可信风险，并构建适配古籍特性的涵盖基础层、内容层、知识层与服务层的可信度评估框架；然后，据此构建包含13项评估指标的GenAI赋能古籍数字化可信度评估指标体系，实现对生成内容的结构化、可操作的可信度评估；最后，提出建立分层响应机制、加强资源分级管理与权责协同的制度保障、构建可信治理长效机制三个方面的可信治理策略，形成“可信风险分析—可信度评估—可信治理”的完整闭环，旨在为GenAI在古籍数字化过程中的合规化、高质量的应用提供理论与实践参考。 |
| 英文摘要: |
| While Generative Artificial Intelligence (GenAI) empowers the digitalization of ancient texts in public libraries, it also introduces trustworthiness risks such as data distortion, model hallucination, and ambiguity of rights and responsibilities, which urgently require systematic governance. This study first analyzes the trustworthiness risks arising from the application of GenAI in the digitalization of ancient texts, and then constructs a four-layer trustworthiness assessment framework tailored to the characteristics of ancient texts, covering the foundation layer, content layer, knowledge layer, and service layer. Based on this framework, a trustworthiness assessment system comprising 13 evaluation metrics is designed to enable structured and operable judgment of generated content. Finally, three trust governance strategies are proposed: establishing a tiered response mechanism, strengthening institutional safeguards for graded
resource management and collaborative responsibility allocation, and building a long-term trust governance mechanism, thereby forming a complete closed loop of trust risk analysis—trustworthiness assessment—trust governance. The aim is to provide theoretical and practical references for the compliant and highquality application of GenAI in the digitalization of ancient texts. |
|
查看全文
查看/发表评论 下载PDF阅读器 |
| 关闭 |
