---
task_categories:
- text-classification
language:
- zh
license: cc-by-sa-4.0
---
# Chinese Classification Dataset

## Overview
The Chinese Classification Dataset is designed for classifying sentences in different forms of Chinese text. Each entry consists of a Chinese sentence paired with a label indicating its language variant.

## Structure
- **Rows**: Each row contains a single Chinese sentence.
- **Labels**: Comma-separated strings indicating one or more of these languages:
  - `zh`: Simplified Chinese
  - `zht`: Traditional Chinese
  - `yue`: Cantonese

### Example
| Sentence                     | Label           |
|------------------------------|-----------------|
| 这是一个示例句子。               | zh              |
| 這是一個示例句子。             | zht             |
| 呢個係一個示例句子。           | yue             |

## Purpose
This dataset contains the same information as [agentlans/cantonese-chinese](https://huggingface.co/datasets/agentlans/cantonese-chinese)
but it's reformulated as a multi-label language classification problem.

## Applications
- Language classification
- NLP research on Chinese languages
- Language detection algorithms
- Educational tools

## Usage Notes
Ideal for training models to identify and classify various forms of Chinese text, with clear labeling for easy integration into machine learning workflows.

## Licence and Acknowledgements

Please refer to [agentlans/cantonese-chinese](https://huggingface.co/datasets/agentlans/cantonese-chinese) and the datasets that it's based on.