Decoding Traditional Chinese Patterns with the WenYi Structured Corpus
Traditional Chinese patterns serve as a key visual language of Chinese culture, with unique aesthetic and symbolic value. However, the lack of comprehensive digital resources and structured representations has limited both large-scale cultural analysis and the culture-grounded reuse of patterns in contemporary design and generative workflows. In response, we construct WenYi , a structured corpus of traditional Chinese patterns. WenYi contains 2978 pattern instances covering 154 motifs across more than 2000 years of Chinese history, organized through a four-dimensional schema of Context, Content, Visual Form, and Cultural Semantics. The corpus is constructed through a hybrid annotation strategy that combines hierarchical metadata fields, controlled vocabularies, and open-ended semantic descriptions. Users can access the corpus through an explorer system we developed. We demonstrate the utility of WenYi through four cases: motif co-occurrence analysis, visual trend exploration, text-to-image model fine-tuning, and semantic-guided design ideation. The study provides a structured foundation for analyzing and reusing traditional Chinese patterns, while also discussing insights into structuring cultural complexity and making visual culture data explicit.
Authors
- Ziyao Gao (ORCID: https://orcid.org/0009-0007-2424-0796)
- Wei Zeng (ORCID: https://orcid.org/0000-0002-5600-8824)
- Huangjing Zhou (ORCID: https://orcid.org/0009-0001-0373-9893)
Institutions
- China University of Geosciences (CN)
- The Hong Kong University of Science and Technology (Guangzhou) (CN)
Publication Details
- Journal
- Journal on Computing and Cultural Heritage
- Published
- 2026-11-17
- DOI
- https://doi.org/10.1145/3862860
- Primary Topic
- Digital Humanities and Scholarship
- Type
- article
- Field-Weighted Citation Impact
- 0.00