An AI model for discovering allusions in Classical Chinese texts
Duoshi is a lightweight and intuitive AI tool designed to help you:
- 🔍 Detect allusions in classical Chinese poetry and prose
- 🧠 Understand hidden meanings behind historical references
- 🔗 Explore intertextual connections across texts
Built on structured knowledge from traditional Chinese leishu類書 (encyclopedic compendia), Duoshi brings together classical scholarship and modern AI.
Developed by:
- Wang Zhitian — Xiamen University Malaysia
- Zhang Jie — Southwest Jiaotong University
Reading classical Chinese texts often requires deep familiarity with allusions.
Duoshi helps you:
- Save time on manual lookup
- Discover connections you might miss
- Support research, teaching, and learning
Text → Find allusions → Link meanings → Generate insights
That’s it. No complicated setup needed.
- 📖 Classical poetry analysis
- 🎓 Teaching Chinese literature
- 🧑🔬 Digital humanities research
- 🤖 AI-assisted annotation tools
git clone https://github.com/ketsukun/Duoshi.git
cd Duoshi
pip install -r requirements.txtfrom duoshi import DuoshiModel
model = DuoshiModel.load("duoshi-base")
text = "青松架短壑，赤脚踏层冰"
result = model.analyze(text)
print(result.allusions)- More leishu datasets
- Better semantic understanding
- Visualization tools
- Integration with more AI systems
If you find this project useful, please cite:
Wang Zhitian & Zhang Jie Duoshi: An Allusion Recognition Model for Classical Chinese Texts
If you like this project, give it a ⭐ on GitHub!
MIT License