File size: 1,596 Bytes
a3824ba
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
19c7328
a3824ba
 
 
 
 
 
 
 
 
 
19c7328
a3824ba
19c7328
a3824ba
 
 
 
d3d28f0
a3824ba
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
---
language:
- id
tags:
- punctuation prediction
- punctuation
widget:
- text: "halo bagaimana kabarmu"
  example_title: "indonesian"
---

This model predicts the punctuation of Indonesian languange. It has been created to restore punctuation of transcribed from speech recognition models.
This model Based on the work https://github.com/oliverguhr/fullstop-deep-punctuation-prediction

The model restores the following punctuation markers: **"." "," "?" "-" ":"**

## Install 

To get started install the package from [pypi](https://pypi.org/project/deepmultilingualpunctuation/):

```bash
pip install deepmultilingualpunctuation
```
### Restore Punctuation
```python
from deepmultilingualpunctuation import PunctuationModel

model = PunctuationModel("Rizkinoor16/fullstop-indonesian-punctuation-prediction")
text = "halo bagaimana kabarmu"
result = model.restore_punctuation(text)
print(result)
```


## Results
```
      precision    recall  f1-score   support 

           0       0.98      0.99      0.98  38057720
           .       0.89      0.91      0.90   2234980
           ,       0.84      0.79      0.81   3037655
           ?       0.84      0.79      0.82     72969
           -       0.96      0.90      0.93    162085
           :       0.91      0.89      0.90    191937

    accuracy                           0.97  43757346
   macro avg       0.90      0.88      0.89  43757346  
weighted avg       0.97      0.97      0.97  43757346
```

## Contact

Rizki Noor <rizki@cakra.ai>

**Linkedin** : [Noor Muhamad Rizki](https://www.linkedin.com/in/noor-muhamad-rizki-114600231/)