---
dataset_info:
  features:
    - name: id
      dtype: string
    - name: meter_id
      dtype: int64
    - name: meter_name
      dtype: string
    - name: verse
      dtype: string
    - name: verse_part1
      dtype: string
    - name: verse_part2
      dtype: string
  splits:
    - name: train
      num_bytes: 6949029
      num_examples: 47124
    - name: test
      num_bytes: 1255648
      num_examples: 8316
  download_size: 6429814
  dataset_size: 8204677
configs:
  - config_name: default
    data_files:
      - split: train
        path: train-*.parquet
      - split: test
        path: test-*.parquet
---

# Arabic Poetry Meter Classification

This dataset is for Arabic poetry meter classification. The dataset contains Arabic verses and their corresponding meter classes.

## Dataset Information

The dataset is divided into training and testing sets:
- **Training set**: More than 47,124 verses
- **Test set**: More than 8,316 verses
- **Total**: 55,440 verses

## Meter Classes

The dataset includes 14 meter classes (0-13):

| Meter ID | Meter Name |
|----------|-----------|
| 0 | Saree |
| 1 | Kamel |
| 2 | Mutakareb |
| 3 | Mutadarak |
| 4 | Munsareh |
| 5 | Madeed |
| 6 | Mujtath |
| 7 | Ramal |
| 8 | Baseet |
| 9 | Khafeef |
| 10 | Taweel |
| 11 | Wafer |
| 12 | Hazaj |
| 13 | Rajaz |

## Data Format

The dataset is stored in Parquet format with the following fields:

- `id`: Unique identifier
- `meter_id`: Meter class ID (0-13)
- `meter_name`: Meter class name
- `verse`: Full verse text (with '#' separator removed)
- `verse_part1`: First part of the verse (before '#')
- `verse_part2`: Second part of the verse (after '#')

## Original Data Format

In the original files (train.txt and test.txt), each row contains:
- Meter ID and verse separated by a space
- Verse consists of two parts separated by '#'

Example: `1 يا ضيفَ طيفٍ ما هَداهُ لمَضجَعي # إلا لهيبٌ في الحشى يتوقّدُ`

## Citation

If you use this dataset, please cite:

```
@article{ALSHAIBANI2020106497,
  title = {MetRec: A dataset for meter classification of arabic poetry},
  journal = {Data in Brief},
  volume = {33},
  pages = {106497},
  year = {2020},
  doi = {https://doi.org/10.1016/j.dib.2020.106497},
  author = {Maged S. Al-shaibani and Zaid Alyafeai and Irfan Ahmad},
  keywords = {Arabic, meter, poetry, classification, prosody}
}
```

## Source

The original dataset was collected from the Aldiwan website. For more information, please refer to the original paper.