面向人工智能的博物馆藏品知识组织——以故宫博物院“中国古代可移动文物概念参考模型”为例

On AI-Oriented Museum-Collection Knowledge System Construction with‘The Conceptual Reference Model of Chinese Ancient Movable Cultural Heritages(CRM-ACA)’Program of the Palace Museum

本文概述了大数据时代中国博物馆在藏品知识组织体系构建和共享方面面临的挑战,诸如文物名称复杂,信息结构化和标准化程度低,相关概念间缺乏有效关系模型等。上述问题降低了藏品信息的机器可读性,给规模化利用藏品数据集进行数字人文研究带来困难。基于此,故宫博物院构建了“中国古代可移动文物概念参考模型”(CRMACA),在专家编目基础上,内化领域知识体系、增强语义网络并构建知识图谱。依托故宫博物院体系完备、品类丰富的藏品,该模型为组织和利用中国古代可移动文物知识提供了一个可扩展的智能化框架,其对齐开放数据标准,为研究和利用释放了新的潜力。

In view of the challenging complicated artifacts'titles,low-level information structuralization and standardization and the absence of effective-relationship models between the related concepts Chinese museums are confronted with in constructing and sharing collection knowledge system in the era of big data,this thesis takes the Palace Museum's program of‘The Conceptual Reference Model of Chinese Ancient Movable Cultural Heritages(CRM-ACA)’as a case to discuss internalizing knowledge system,enhancing semantic network and constructing knowledge graph,etc.The CRM-ACA program takes the advantage of the treasures of the Palace Museum to make an extensible intelligent framework that will contribute to organizing and utilizing the knowledge about Chinese ancient movable cultural heritages aligning with the open data criteria to make more sources available for researches and uses.