Datasets:
Indonesian Online Gambling Promotion Multimodal Dataset
This dataset is intended for machine learning research on detecting and filtering online gambling promotions on Indonesian social media. It is a research artifact for content moderation and harmful content detection, not a gambling website or promotional resource.
Research Context
This dataset is created for academic research in harmful content detection and multimodal machine learning. It is intended to help platforms and researchers develop AI systems to identify and remove gambling spam β not to promote gambling.
- Research Paper: Multimodal Deep Learning for Online Gambling Detection on Social Media in Indonesia
- License: CC BY 4.0 (Academic use)
- Institution: Institut Teknologi Sepuluh Nopember
Dataset Purpose
- Purpose: Train machine learning models to identify gambling promotions versus legitimate content
- Content: 5,028 annotated social media posts (gambling spam samples + non-gambling samples)
- Use Case: Binary classification for content moderation systems
- Precedent: Similar to spam detection datasets, phishing email datasets, malware analysis datasets β standard in cybersecurity and NLP research
Scope and Restrictions
- This is not a gambling website.
- This is not promotional content for gambling.
- This is not intended for commercial gambling operations.
- This dataset must not be used to encourage, facilitate, or promote gambling.
A multimodal dataset of 5,028 annotated social media posts for detecting online gambling promotions in Indonesia, containing paired images and Indonesian text captions collected from Facebook, TikTok, and X.
Dataset Statistics
| Platform | Gambling | Non-Gambling | Total |
|---|---|---|---|
| 927 | 492 | 1,419 | |
| TikTok | 199 | 1,243 | 1,442 |
| X | 1,069 | 1,098 | 2,167 |
| Total | 2,195 (43.7%) | 2,833 (56.3%) | 5,028 |
Data Format
data/
βββ dataset.csv # Annotations and metadata
βββ image/ # 5,028 images
βββ image_1.jpg
βββ image_2.jpg
βββ ...
Columns in dataset.csv:
| Column | Type | Description |
|---|---|---|
id |
int | Unique sample ID |
text_caption |
str | Indonesian text caption from the post |
image_path |
str | Relative path to the image file (e.g., image/image_1.jpg) |
platform |
str | Source platform: Facebook, TikTok, or X |
label |
float | 1.0 = gambling promotion, 0.0 = non-gambling |
Collection Method
Posts were collected from Facebook, TikTok, and X using various gambling-related keywords for research purposes only. After deduplication and filtering, the authors manually annotated each sample as gambling promotion (1) or non-gambling (0).
Ethics Statement
- All data was collected from publicly available sources
- No private or personal data was intentionally collected
- The dataset is intended solely for academic and research purposes in content moderation
- This dataset follows standard practices for research datasets in machine learning and cybersecurity
Intended Use
This dataset is intended for:
- Machine learning research
- Content moderation system development
- Harmful content detection research
This dataset is NOT intended for:
- Promoting or facilitating gambling
- Commercial gambling operations
- Redistributing gambling content for promotional purposes
Citation
If you use this dataset in your research, please cite:
Ekaputra, D. R. (2026). Indonesian Online Gambling Promotion Multimodal Dataset [Data set]. Hugging Face. https://huggingface.co/datasets/0xRafie/indonesian-gambling-multimodal-dataset
License
This dataset is released under the CC BY 4.0 license for academic and research use.
Note: This license does not grant permission to use the dataset for illegal activities or gambling promotion.
Disclaimer
This dataset contains a research artifact for machine learning classification of harmful content. It is analogous to phishing detection datasets, spam classification datasets, and malware analysis datasets commonly hosted for academic research. The dataset is not intended to promote gambling but to help detect and filter gambling spam on social media platforms.
- Downloads last month
- 7,220