Update README.md
Browse files
README.md
CHANGED
|
@@ -1,3 +1,53 @@
|
|
| 1 |
-
---
|
| 2 |
-
license: cc-by-4.0
|
| 3 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
---
|
| 2 |
+
license: cc-by-4.0
|
| 3 |
+
language:
|
| 4 |
+
- zh
|
| 5 |
+
- ug
|
| 6 |
+
- bo
|
| 7 |
+
- en
|
| 8 |
+
---
|
| 9 |
+
|
| 10 |
+
# CUTE Dataset
|
| 11 |
+
|
| 12 |
+
CUTE (Chinese, Uyghur, Tibetan, English) 是一个大规模多语言数据集，专门设计用于增强低资源语言的跨语言知识迁移。数据集包含平行语料和非平行语料两部分，总规模约50GB。
|
| 13 |
+
|
| 14 |
+
## 数据集组成
|
| 15 |
+
|
| 16 |
+
### 平行语料 (24.70GB)
|
| 17 |
+
- 中文：2.62GB
|
| 18 |
+
- 英语：3.49GB
|
| 19 |
+
- 维吾尔语：7.37GB
|
| 20 |
+
- 藏语：11.22GB
|
| 21 |
+
|
| 22 |
+
### 非平行语料 (25.80GB)
|
| 23 |
+
- 中文：2.64GB
|
| 24 |
+
- 英语：3.49GB
|
| 25 |
+
- 维吾尔语：7.77GB
|
| 26 |
+
- 藏语：11.90GB
|
| 27 |
+
|
| 28 |
+
## 数据质量
|
| 29 |
+
|
| 30 |
+
数据集通过机器翻译生成，并经过人工评估验证：
|
| 31 |
+
- 中英翻译平均得分：9.1
|
| 32 |
+
- 中维翻译平均得分：8.5
|
| 33 |
+
- 中藏翻译平均得分：8.6
|
| 34 |
+
|
| 35 |
+
## 相关链接
|
| 36 |
+
|
| 37 |
+
- 📝 [论文](https://aclanthology.org/2025.coling-main.670/)
|
| 38 |
+
- 🤖 [模型](https://huggingface.co/CMLI-NLP/CUTE-Llama)
|
| 39 |
+
- 📂 [GitHub](https://github.com/CMLI-NLP/CUTE)
|
| 40 |
+
|
| 41 |
+
## 引用
|
| 42 |
+
|
| 43 |
+
如果您使用了我们的数据集，请引用我们的论文：
|
| 44 |
+
|
| 45 |
+
```bibtex
|
| 46 |
+
@inproceedings{zhuang2025cute,
|
| 47 |
+
title={CUTE: A Multilingual Dataset for Enhancing Cross-Lingual Knowledge Transfer in Low-Resource Languages},
|
| 48 |
+
author={Zhuang, Wenhao and Sun, Yuan},
|
| 49 |
+
booktitle={Proceedings of the 31st International Conference on Computational Linguistics},
|
| 50 |
+
pages={10037--10046},
|
| 51 |
+
year={2025}
|
| 52 |
+
}
|
| 53 |
+
```
|