Datasets:

chuuhtetnaing
/

dhamma-article-dataset

dataset_info:
  features:
    - name: url
      dtype: string
    - name: title
      dtype: string
    - name: body
      dtype: string
  splits:
    - name: train
      num_bytes: 15006551
      num_examples: 1434
  download_size: 4842555
  dataset_size: 15006551
configs:
  - config_name: default
    data_files:
      - split: train
        path: data/train-*

Dhamma Article Dataset (Last Crawl Date: 18/04/2025)

This dataset contains Dhamma articles scraped from the Dhamma Ransi website.

Dataset Description

Overview

The dataset consists of Dhamma articles collected from the Dhamma Ransi website, providing a valuable resource for Myanmar language processing tasks and Buddhist text analysis.

Data Collection Methodology

The data was ethically scraped with:

Adherence to the website's robots.txt guidelines
3-second delays between API calls to minimize server load
Collection conducted responsibly to avoid stressing the website

Data Structure

The dataset contains the following fields:

Column	Description
`url`	Source URL where the article was scraped from
`title`	Title of the Dhamma article
`body`	Full text content of the article

Usage Examples

Loading the Dataset

from datasets import load_dataset

# Load the dataset from Hugging Face
ds = load_dataset("chuuhtetnaing/dhamma-article-dataset")

# Access the first example
first_article = ds['train'][0]
print(f"Title: {first_article['title']}")
print(f"Body: {first_article['body']}")

Using for NLP Tasks

# Example of preparing the dataset for text classification
from transformers import AutoTokenizer

tokenizer = AutoTokenizer.from_pretrained("your-tokenizer-model")

def tokenize_function(examples):
    return tokenizer(examples["body"], padding="max_length", truncation=True, max_length=512)

tokenized_ds = ds.map(tokenize_function, batched=True)

Dataset Purpose

This dataset is designed for multiple Myanmar language NLP tasks including:

Masked Language Modeling (MLM)
Next Sentence Prediction (NSP)
Other self-supervised pretraining tasks
General Myanmar language model development

Limitations

This dataset contains copyrighted material and should be used for research purposes only.
The articles may have varying lengths and formats.
The dataset represents content available on the source website as of the last crawl date.

Message from Dataset Creator

This dataset has been created for research purposes to advance natural language processing and machine learning capabilities for the Myanmar language, particularly focusing on Buddhist texts and Dhamma teachings. We respectfully request that users of this dataset:

Use this collection for academic, research, and educational purposes only
Acknowledge the original authors and publishers whose works appear in this corpus
Respect copyright considerations and intellectual property rights
Consider supporting the Dhamma Ransi organization by visiting their website
Contribute to the development of digital resources for Myanmar language preservation and Buddhist text analysis

The creation of this dataset is intended to bridge the gap in language resources for Myanmar NLP research while respecting the importance of supporting Dhamma teachings and Buddhist scholarship. We kindly ask all researchers to use this data responsibly.

Acknowledgments

This dataset is derived from the Dhamma Ransi website.