Update README.md
Browse files
README.md
CHANGED
|
@@ -1,3 +1,4 @@
|
|
|
|
|
| 1 |
---
|
| 2 |
language:
|
| 3 |
- ar
|
|
@@ -39,11 +40,8 @@ model-index:
|
|
| 39 |
- Damir Cavar
|
| 40 |
conference: "FinNLP / FNP / LLMFinLegal Workshop (2025)"
|
| 41 |
year: 2025
|
| 42 |
-
url: https://aclanthology.org/2025.finnlp-1.20
|
| 43 |
---
|
| 44 |
-
|
| 45 |
-
|
| 46 |
-
|
| 47 |
```
|
| 48 |
|
| 49 |
# AMWAL: Arabic Financial Named Entity Recognition (NER)
|
|
@@ -53,7 +51,7 @@ model-index:
|
|
| 53 |
### Install (recommended)
|
| 54 |
|
| 55 |
```bash
|
| 56 |
-
pip install git+https://huggingface.co/Muhsabrys/
|
| 57 |
```
|
| 58 |
|
| 59 |
```python
|
|
@@ -166,26 +164,20 @@ The model recognizes **21 financial entity types**, including (but not limited t
|
|
| 166 |
|
| 167 |
## Evaluation Results
|
| 168 |
|
| 169 |
-
The model was evaluated on a held-out test set using standard NER metrics:
|
| 170 |
-
|
| 171 |
| Metric | Score |
|
| 172 |
| --------- | ---------- |
|
| 173 |
| Precision | **96.08%** |
|
| 174 |
| Recall | **95.87%** |
|
| 175 |
| F1-score | **95.97%** |
|
| 176 |
|
| 177 |
-
These results are competitive with reported financial NER systems in other languages, despite the additional challenges posed by Arabic morphology and orthography.
|
| 178 |
-
|
| 179 |
---
|
| 180 |
|
| 181 |
## Usage
|
| 182 |
|
| 183 |
-
AMWAL supports **two officially supported usage modes**.
|
| 184 |
-
|
| 185 |
### Option 1 โ Install via `pip` (recommended)
|
| 186 |
|
| 187 |
```bash
|
| 188 |
-
pip install git+https://huggingface.co/Muhsabrys/
|
| 189 |
```
|
| 190 |
|
| 191 |
```python
|
|
@@ -203,7 +195,7 @@ result = ner("ูุต ุนุฑุจู ู
ุงูู")
|
|
| 203 |
from huggingface_hub import snapshot_download
|
| 204 |
import sys
|
| 205 |
|
| 206 |
-
repo_path = snapshot_download("Muhsabrys/")
|
| 207 |
sys.path.append(repo_path)
|
| 208 |
|
| 209 |
from amwal import load_ner
|
|
@@ -244,8 +236,6 @@ result = ner("ูุต ุนุฑุจู ู
ุงูู")
|
|
| 244 |
|
| 245 |
## Future Work
|
| 246 |
|
| 247 |
-
Planned future directions include:
|
| 248 |
-
|
| 249 |
* Expanding the annotated corpus
|
| 250 |
* Introducing hierarchical entity structures
|
| 251 |
* Modeling relations between financial entities
|
|
@@ -263,4 +253,4 @@ Planned future directions include:
|
|
| 263 |
pages={207--213},
|
| 264 |
year={2025}
|
| 265 |
}
|
| 266 |
-
```
|
|
|
|
| 1 |
+
```yaml
|
| 2 |
---
|
| 3 |
language:
|
| 4 |
- ar
|
|
|
|
| 40 |
- Damir Cavar
|
| 41 |
conference: "FinNLP / FNP / LLMFinLegal Workshop (2025)"
|
| 42 |
year: 2025
|
| 43 |
+
url: https://aclanthology.org/2025.finnlp-1.20
|
| 44 |
---
|
|
|
|
|
|
|
|
|
|
| 45 |
```
|
| 46 |
|
| 47 |
# AMWAL: Arabic Financial Named Entity Recognition (NER)
|
|
|
|
| 51 |
### Install (recommended)
|
| 52 |
|
| 53 |
```bash
|
| 54 |
+
pip install git+https://huggingface.co/Muhsabrys/AMWAL-ner-arabic
|
| 55 |
```
|
| 56 |
|
| 57 |
```python
|
|
|
|
| 164 |
|
| 165 |
## Evaluation Results
|
| 166 |
|
|
|
|
|
|
|
| 167 |
| Metric | Score |
|
| 168 |
| --------- | ---------- |
|
| 169 |
| Precision | **96.08%** |
|
| 170 |
| Recall | **95.87%** |
|
| 171 |
| F1-score | **95.97%** |
|
| 172 |
|
|
|
|
|
|
|
| 173 |
---
|
| 174 |
|
| 175 |
## Usage
|
| 176 |
|
|
|
|
|
|
|
| 177 |
### Option 1 โ Install via `pip` (recommended)
|
| 178 |
|
| 179 |
```bash
|
| 180 |
+
pip install git+https://huggingface.co/Muhsabrys/AMWAL-ner-arabic
|
| 181 |
```
|
| 182 |
|
| 183 |
```python
|
|
|
|
| 195 |
from huggingface_hub import snapshot_download
|
| 196 |
import sys
|
| 197 |
|
| 198 |
+
repo_path = snapshot_download("Muhsabrys/AMWAL-ner-arabic")
|
| 199 |
sys.path.append(repo_path)
|
| 200 |
|
| 201 |
from amwal import load_ner
|
|
|
|
| 236 |
|
| 237 |
## Future Work
|
| 238 |
|
|
|
|
|
|
|
| 239 |
* Expanding the annotated corpus
|
| 240 |
* Introducing hierarchical entity structures
|
| 241 |
* Modeling relations between financial entities
|
|
|
|
| 253 |
pages={207--213},
|
| 254 |
year={2025}
|
| 255 |
}
|
| 256 |
+
```
|