University of Illinois Urbana-Champaign

Genai-enabled construction for a hierarchical descriptive model of ancient painting resources - A case study on Chinese paintings in the palace museum

Li, Jiayi; Sun, Zhihan; Lin, Duo; Yang, Xuan; Liu, Hengchang; Yan, Chenxi

Permalink

https://hdl.handle.net/2142/126225

Description

Title
Genai-enabled construction for a hierarchical descriptive model of ancient painting resources - A case study on Chinese paintings in the palace museum
Author(s)
  • Li, Jiayi
  • Sun, Zhihan
  • Lin, Duo
  • Yang, Xuan
  • Liu, Hengchang
  • Yan, Chenxi
Issue Date
2025-03-11
Keyword(s)
  • Ancient Chinese painting
  • Hierarchical descriptive model
  • Semantic organization
  • Large language models
  • Semantic queries
Abstract
The digital research of ancient Chinese paintings plays an important role in the field of cultural heritage, but the existing research on the organization of ancient painting resources lacks in-depth excavation of the artistic connotations of the paintings and consideration of the historical impact and audience perception. Therefore, based on the Panofsky-Shatford model, this paper combined Xie He’s Six Principles in Painting and audience response theory to construct a hierarchical descriptive model (HDM) that includes three dimensions: painting content, cultural effect and metadata. Meanwhile, we realized an efficient knowledge extraction workflow with the Generative Artificial Intelligence technology (i.e. large language models and prompt engineering), and evaluated the effectiveness of HDM by testifying the advantageous performance its application on semantic querying in graph databases. The experiment results on a collected painting dataset of The Palace Museum show that HDM not only can relates and integrates the basic metadata and content knowledge of paintings in a systematic way, but also incorporates diverse viewer insights and historical meanings, thus deepening the user's knowledge learning and sense making process.
Publisher
iSchools
Series/Report Name or Number
iConference 2025 Proceedings
Type of Resource
Other
Genre of Resource
Conference Poster
Language
eng
Handle URL
https://hdl.handle.net/2142/126225
Copyright and License Information
Copyright 2025 is held by Jiayi Li, Zhihan Sun, Duo Lin, Xuan Yang, Hengchang Liu, and Chenxi Yan. Copyright permissions, when appropriate, must be obtained directly from the author.

Owning Collections

iConference 2025 Posters PRIMARY
Posters presented at the 2025 iConference https://www.ischools.org/iconference