Datasets:
Tasks:
Token Classification
Modalities:
Text
Formats:
parquet
Languages:
Arabic
Size:
< 1K
Tags:
segmentation
License:
Update README.md
Browse files
README.md
CHANGED
|
@@ -49,6 +49,8 @@ The corpus is designed to support research on sentence segmentation in Modern St
|
|
| 49 |
AraSeg contains manually annotated documents collected from diverse sources and genres, enabling robust evaluation across different writing styles and domains.
|
| 50 |
The benchmark contains manually annotated documents collected from diverse sources and genres, enabling robust evaluation across different writing styles and domains.
|
| 51 |
|
|
|
|
|
|
|
| 52 |
---
|
| 53 |
## Dataset Structure
|
| 54 |
|
|
|
|
| 49 |
AraSeg contains manually annotated documents collected from diverse sources and genres, enabling robust evaluation across different writing styles and domains.
|
| 50 |
The benchmark contains manually annotated documents collected from diverse sources and genres, enabling robust evaluation across different writing styles and domains.
|
| 51 |
|
| 52 |
+
AraSeg-PA is the Paragraph-Aware (PA) variant of the corpus where paragraph boundaries are provided.
|
| 53 |
+
|
| 54 |
---
|
| 55 |
## Dataset Structure
|
| 56 |
|