Palatial studies
Art historyIconographyBiology
The “Ancient Chinese Artifacts Knowledge Graph” (KG-ACA) is a long-term data enrichment project initiated by the Palace Museum in 2019, aimed at enhancing public cultural services. Building on the museum’s collection cataloging information, the project expands to include interdisciplinary concepts and terminologies from fields such as palatial studies, art history, iconography, and biology. This allows users to find and utilize the museum's collection information from multiple dimensions.
Since the mid-1990s, with the comprehensive development of digitization and informatization in the cultural heritage sector, global cultural heritage institutions have released a vast amount of collection information through the internet, including basic information, images, videos, and 3D models. These artifacts, once hidden in storage, now have unprecedented opportunities to digitally meet users worldwide. Every year, millions of searches are conducted by users in the Palace Museum's Collection Online Platform (Shuzi wenwuku) to find specific collections of certain subject matters. Users typically employ common, everyday language in their searches. However, the collection information in the Collection Online Platform is directly sourced from the museum’s collection management system, where curators record known information about artifacts using standardized and highly specialized terminology. This discrepancy in language can lead to incomplete search results.
To tackle this problem, the Palace Museum has explored the use of knowledge graph technology from the perspectives of computer science and information resource management. This aims to overcome the limitations of exact keyword matching in traditional search, making it easier to find artifacts in the Collection Online Platform.
In 2020, with funding from the Ministry of Culture and Tourism's “Research Project on Culture, Art and Tourism—Information Technology Development Project,” the Palace Museum initiated research on the “Conceptual Reference Model for Movable Cultural Relics Applied to Smart Search.” Following international standards (ISO 21127:2023) and national standards (GB/T 37965-2019), the project constructed the “Ancient Chinese Artifacts Conceptual Reference Model” (CRM-ACA) as the ontology for the knowledge graph. Concurrently, a comprehensive thesaurus containing over 52,000 terms was developed, covering dimensions such as object type, motif and pattern, color, form and structure, process and technique, subject matter, and person. This supports continuous human-machine collaborative annotation and data enrichment of collection information.
This data visualization showcases a “capillary” of the knowledge graph, presenting the relationships between the annotated artifacts and related concepts within the Collection Online Platform. This semantic network continues to grow, allowing for limitless expansion. Users can click on any node to explore multi-dimensional,multi-layered connections among these artifacts, uncovering previously unnoticed relationships and similarities. As the project progresses, the Palace Museum will further extend collection information services based on the knowledge graph to more smart use scenarios. We welcome feedback and suggestions from all parties and invite colleagues from cultural heritage institutions with similar needs to join us in sharing and building this knowledge graph.
Contact us: knowledgegraph@dpm.org.cn
Experts from various fields at the Palace Museum provided significant support in constructing the "Ancient Chinese Artifacts Thesaurus." Vice Directors Ren Wanping and Zhu Hongwen coordinated expert consultations and reviews. Contributing experts include (in alphabetical order): Da Weijia, Guo Fuxiang, Han Qian, Jin Lu, Lyv Chenglong, Wang Qi, Wang Hao, Wang Yi, Wen Ming, Xu Haifeng, Yan Yong, Zeng Jun, Zhang Nan. Reviewing experts include: Da Weijia, Han Qian, Ji Luoyuan, Liu Yue, Wang He, Wu Qian, Xu Haifeng, Yu Wentao, Zhang Linjie, Zhang Xin, Zhao Yingying.
Dr. Martin Doerr, a member of the CIDOC CRM SIG (Special Interest Group), provided substantial support during CRM-ACA ontology building process, with his feedback and suggestions greatly inspiring the ontology development.
Professor Ren Ming of Renmin University of China, Professor Huang Mingyu of University of Science and Technology Beijing, Professor Wang Yujue of Wuhan University, and Senior Engineer Chen Can of Computer Network Information Center of Chinese Academy of Sciences offered valuable insights and suggestions when the project was in progress.
Victoria and Albert Museum's “ Chinese Iconography Thesaurus ” team generously shared their thematic thesaurus construction outcomes with our project team.
Language