Update README.md
Browse files
README.md
CHANGED
|
@@ -1,3 +1,71 @@
|
|
| 1 |
-
---
|
| 2 |
-
license: cc-by-4.0
|
| 3 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
---
|
| 2 |
+
license: cc-by-4.0
|
| 3 |
+
language: ar
|
| 4 |
+
tags:
|
| 5 |
+
- named-entity-recognition
|
| 6 |
+
- arabic
|
| 7 |
+
- human-annotated
|
| 8 |
+
- iahlt
|
| 9 |
+
dataset_name: iahlt-ner-arabic
|
| 10 |
+
---
|
| 11 |
+
|
| 12 |
+
# IAHLT Named Entities Dataset (Arabic Subset)
|
| 13 |
+
|
| 14 |
+
**האיגוד הישראלי לטכנולוגיות שפת אנוש**
|
| 15 |
+
**الرابطة الإسرائيلية لتكنولوجيا اللغة البشرية**
|
| 16 |
+
**The Israeli Association of Human Language Technologies**
|
| 17 |
+
https://www.iahlt.org
|
| 18 |
+
|
| 19 |
+
This dataset contains named entity annotations for Arabic texts from various sources, curated as part of the IAHLT multilingual NER project. The Arabic portion is provided here as a cleaned subset intended for training and evaluation in named entity recognition tasks.
|
| 20 |
+
|
| 21 |
+
## Files Included
|
| 22 |
+
|
| 23 |
+
This release includes the following JSONL files:
|
| 24 |
+
|
| 25 |
+
- `iahlt_ner_train.jsonl`
|
| 26 |
+
- `iahlt_ner_val.jsonl`
|
| 27 |
+
- `iahlt_ner_test.jsonl`
|
| 28 |
+
|
| 29 |
+
Each file contains one JSON object per line with the following fields:
|
| 30 |
+
- `text`: the raw paragraph text
|
| 31 |
+
- `label`: a list of triples `[start, end, label]` where:
|
| 32 |
+
- `start` is the index of the first character of the entity
|
| 33 |
+
- `end` is the index of the first character after the entity
|
| 34 |
+
- `label` is the entity class (e.g., `PER`, `ORG`, etc.)
|
| 35 |
+
- `metadata`: a dictionary with source and annotation metadata
|
| 36 |
+
|
| 37 |
+
## Entity Types
|
| 38 |
+
|
| 39 |
+
The dataset includes the following entity types (summed across splits):
|
| 40 |
+
|
| 41 |
+
| Entity | Count |
|
| 42 |
+
|-----------|-------|
|
| 43 |
+
| GPE | 26,767 |
|
| 44 |
+
| PER | 22,694 |
|
| 45 |
+
| ORG | 19,906 |
|
| 46 |
+
| TIMEX | 10,288 |
|
| 47 |
+
| TTL | 10,075 |
|
| 48 |
+
| FAC | 4,740 |
|
| 49 |
+
| MISC | 7,087 |
|
| 50 |
+
| LOC | 5,389 |
|
| 51 |
+
| EVE | 2,595 |
|
| 52 |
+
| WOA | 1,781 |
|
| 53 |
+
| DUC | 1,715 |
|
| 54 |
+
| ANG | 732 |
|
| 55 |
+
| INFORMAL | 10 |
|
| 56 |
+
|
| 57 |
+
> Note: These statistics reflect a cleaned version of the dataset. Some entities and texts have been modified or removed for consistency and usability.
|
| 58 |
+
|
| 59 |
+
## Annotation Notes
|
| 60 |
+
|
| 61 |
+
- Entity spans were manually annotated at the grapheme level and then normalized using Arabic-specific punctuation and spacing rules.
|
| 62 |
+
- Nested spans are allowed.
|
| 63 |
+
- The dataset was cleaned to ensure compatibility with common NER training formats.
|
| 64 |
+
|
| 65 |
+
## License
|
| 66 |
+
|
| 67 |
+
This dataset is released under the [Creative Commons Attribution 4.0 International](https://creativecommons.org/licenses/by/4.0/) license.
|
| 68 |
+
|
| 69 |
+
## Acknowledgments
|
| 70 |
+
|
| 71 |
+
We thank all annotators and contributors who worked on this corpus.
|