Vietnamese News Dataset

A dataset of Vietnamese news articles collected from 6 major Vietnamese newspapers for NLP research.

Dataset Description

• Homepage: https://github.com/undertheseanlp/underthesea
• Repository: https://huggingface.co/datasets/undertheseanlp/UVN-1
• Point of Contact: Vu Anh (anhv.ict91@gmail.com)

Dataset Summary

This dataset contains 3,268 Vietnamese news articles covering various topics including politics, business, sports, entertainment, education, health, and technology. It is designed for Vietnamese NLP research tasks such as:

• Text classification (news categorization)
• Language modeling
• Text generation
• Named entity recognition
• Keyword extraction

Languages

Vietnamese (vi)

Dataset Structure

Data Instance

{
  "id": "article-00001",
  "source": "vnexpress",
  "url": "https://vnexpress.net/example-123.html",
  "category": "Thời sự",
  "title": "Tiêu đề bài báo",
  "content": "Nội dung bài báo...",
  "publish_date": "2026-01-28"
}

Data Fields

Field	Type	Description
id	string	Unique identifier (article-XXXXX)
source	string	News source
url	string	Original article URL
category	string	News category
title	string	Article title
content	string	Full article content
publish_date	string	Publication date (yyyy-mm-dd)

Data Splits

Split	Count
train	3,268

Sources

Source	Website	Count
VnExpress	vnexpress.net	1,000
Dân Trí	dantri.com.vn	976
Thanh Niên	thanhnien.vn	579
Tuổi Trẻ	tuoitre.vn	210
Tiền Phong	tienphong.vn	240
Người Lao Động	nld.com.vn	263

Categories

• Thời sự (Politics)
• Thế giới (World)
• Kinh doanh (Business)
• Giải trí (Entertainment)
• Thể thao (Sports)
• Pháp luật (Law)
• Giáo dục (Education)
• Sức khỏe (Health)
• Đời sống (Lifestyle)
• Khoa học (Science/Technology)

Usage

from datasets import load_dataset

dataset = load_dataset("undertheseanlp/UVN-1")

# Access articles
for article in dataset["train"]:
    print(article["id"], article["title"])
    break

# Filter by source
vnexpress = dataset["train"].filter(lambda x: x["source"] == "vnexpress")

# Filter by category
sports = dataset["train"].filter(lambda x: x["category"] == "Thể thao")

Considerations

Biases

• Reflects editorial perspectives of mainstream Vietnamese news outlets
• May be biased toward urban and national news coverage

Limitations

• Articles from 2025-2026 time period
• Uneven distribution across sources due to website structure differences

License

CC-BY-NC-4.0. For research and educational purposes only.

Citation

@misc{vietnamese_news_dataset,
  title={Vietnamese News Dataset},
  author={Underthesea NLP},
  year={2026},
  publisher={HuggingFace},
  url={https://huggingface.co/datasets/undertheseanlp/UVN-1}
}