File size: 3,298 Bytes
e727f49
 
 
 
 
 
 
 
 
6388061
 
 
 
7d5510a
6388061
7d5510a
6388061
7d5510a
6388061
 
 
7d5510a
 
 
6388061
 
 
7d5510a
 
6388061
 
7d5510a
 
 
6388061
7d5510a
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
6388061
 
 
 
7d5510a
 
 
 
 
 
 
 
8ddf22f
 
7d5510a
 
 
 
 
 
 
 
 
 
8ddf22f
 
 
 
 
 
 
 
 
 
 
7d5510a
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
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
---
license: mit
task_categories:
- text-classification
language:
- mg
size_categories:
- 1K<n<10K
---
# Dataset Card for Vaovao Malagasy Sentiment Corpus (VMSC)

## Dataset Summary

The **Vaovao Malagasy Sentiment Corpus (VMSC)** is the first publicly available, manually annotated sentiment analysis dataset for the Malagasy language (`mg`). It contains **5,041 sentences** extracted from news articles (*vaovao*) published between 2022 and 2023.

Each sentence is labeled with binary sentiment (**Positive** or **Negative**). The dataset was created to address the scarcity of resources for Malagasy NLP and benchmarks. It features high-quality annotations performed by native speakers, achieving an inter-annotator agreement of **Cohen’s Kappa = 0.96**.

This dataset is the official corpus presented in the paper published in *Language Resources and Evaluation* (Springer, 2025).

## Supported Tasks

- **Sentiment Analysis:** Binary classification (Positive vs. Negative).
- **Transfer Learning:** Benchmarking pre-trained models (e.g., XLM-R, mBERT) on Malagasy.
- **Low-Resource NLP:** Serving as a gold-standard evaluation set for low-resource language studies.

## Languages

- **Language:** Malagasy (`mg`)
- **Variant:** Standard Malagasy (official dialect used in media and government).


## Dataset Structure

### Data Instances

An example of a data point looks like this:

```json
{
  "text": "Tena nety ny vokatra azo tamin'ny tetikasa.",
  "label": 1
}
```
### Data Fields
- text: The news sentence (string).
- label: The sentiment label (integer).
  * 0: Negative
  * 1: Positive
    
### Data Splits
The dataset contains a total of 5,041 sentences.

### Dataset Creation
* Source: Malagasy news articles published between 2022 and 2023.
* Annotation: Manually annotated by native speakers.
* Quality Control: Rigorous annotation protocol resulting in a Cohen’s Kappa score of 0.96.
  
### How to Use
You can load this dataset directly using the Hugging Face datasets library:

```python
from datasets import load_dataset

# Load the dataset
dataset = load_dataset("Lo-Renz-O/vaovao_malagasy_sentiment_corpus")

# View the first training example
print(dataset['train'][0])
```

### Citation
If you use this dataset in your research, please cite both the paper (methodology) and the dataset version:
1. Paper citation:
```bibtex
@article{Mamelona2025VMSC,
  author = {Lorenzo Mamelona, Sitraka Ny Aina Raharivelo, Andriamanarivo Tatiana Miarintsoa, Bright Bediako-Kyeremeh & Benjamin Kwapong Osibo}, 
  title = {Towards robust Malagasy NLP: a novel corpus and evaluation framework for sentiment analysis},
  journal = {Language Resources and Evaluation},
  year = {2026},
  publisher = {Springer},
  doi = {10.1007/s10579-025-09900-w},
  url = {https://link.springer.com/article/10.1007/s10579-025-09900-w}
}
```
2. Dataset Citation (DOI):
```bibtex
@misc{lorenzo_mamelona_2025,
  author       = {Lorenzo Mamelona, Sitraka Ny Aina Raharivelo, Andriamanarivo Tatiana Miarintsoa, Bright Bediako-Kyeremeh & Benjamin Kwapong Osibo},
  title        = {vaovao_malagasy_sentiment_corpus},
  year         = {2025},
  url          = {https://huggingface.co/datasets/Lo-Renz-O/vaovao_malagasy_sentiment_corpus},
  doi          = {10.57967/hf/6293},
  publisher    = {Hugging Face}
}
```