TatarNLPWorld/tatar-news-cluster

Tatar News Clustered Dataset

A comprehensive dataset of Tatar language news articles with topic categories, curated by TatarNLPWorld as part of the Tat2Vec project.

📖 Overview

This dataset contains 57,340 Tatar language texts collected from various sources and organized into topic categories. Each entry includes the full content, title, category, source, and publication date. The dataset is designed for text classification, clustering analysis, and various NLP tasks for Tatar language processing.

📊 Dataset Statistics

General Information

Metric	Value
Total Records	57,340
Records with Title	57,340 (100%)
Records with Date	57,340 (100%)
Total Characters	133,540,250
Average Content Length	2,329 characters
Median Content Length	1,138 characters
Content Length Range	50 - 10,000 characters
Date Range	2024 - 2026
Number of Categories	282

Top Categories

Category (Tatar)	Category (English)	Count	Percentage
Җәмгыять	Society	8,882	15.5%
Мәдәният	Culture	6,641	11.6%
Шоу-бизнес	Show Business	6,440	11.2%
Яңалыклар	News	4,854	8.5%
Дин	Religion	2,135	3.7%
Архив	Archive	1,470	2.6%
Тормыш сулышы	Lifestyle	1,150	2.0%
Мәгариф	Education	1,073	1.9%
Гыйбрәт ал	Lessons	1,071	1.9%
Милләт	Nation	1,054	1.8%
Other categories		22,570	39.3%

Total Unique Categories: 282

Source Distribution

Source	Count	Percentage
merged_matbugat_news	27,539	48.0%
beznen_articles	9,990	17.4%
merged_azatliqorg	5,762	10.0%
syuyumbike_news	3,367	5.9%
tatar_inform	2,562	4.5%
Other sources		14.2%

📁 Data Structure

Each record contains the following fields:

Field	Type	Description
content	string	Full article content in Tatar language
title	string	Article title
category	string	Topic category (282 unique categories)
source	string	Original source name
content_length	integer	Length of content in characters
resource	string	Source URL or identifier
date	string	Publication date

🚀 Usage Example

Loading the Dataset

from datasets import load_dataset
import pandas as pd

# Load the dataset
dataset = load_dataset("TatarNLPWorld/tatar-news-cluster")
data = dataset["train"]

# Display first 3 records as a markdown table (truncated content)
first_three = data[:3]
df = pd.DataFrame(first_three)

# Truncate content for display (first 100 chars)
for idx, row in df.iterrows():
    df.at[idx, 'content'] = row['content'][:100] + "..." if len(row['content']) > 100 else row['content']

markdown_table = df[['title', 'content', 'resource', 'date']].to_markdown(index=False)
print(markdown_table)

Expected Output:

title	content	resource	date
Кыстыбый, күзикмәк, якмыш.	Камыр өчен: - 300 мл җылымса су; - 1 чәй калагы тоз; - 3 аш калагы үсемлек мае; - 1 аш калагы майон...	https://alluki.ru/...	2026-01-06T08:20:00
Ана капиталын баланы укытуга ничек файдаланырга?	Русиянең Социаль фондында ана капиталы акчаларын балаларга өстәмә белем бирү һәм укытуга юнәлтү мөм...	https://alluki.ru/...	2026-02-07T08:00:00
Безнең дуслар – олимпиада җиңүчеләре!	13-14 гыйнварда Стәрлетамак шәһәренең 2 нче интернат-лицеенда 9-11 сыйныф укучылары өчен туган телл...	https://alluki.ru/...	2026-01-15T16:00:00

🔬 Applications

This dataset supports multiple NLP tasks:

• Multi-class Text Classification: Topic categorization with 282 classes
• Text Clustering: Document clustering analysis
• Token Classification: Named entity recognition, part-of-speech tagging
• Feature Extraction: Creating embeddings for Tatar language texts
• Sentence Similarity: Comparing semantic similarity between texts
• Text Generation: Language modeling and text generation tasks
• Summarization: Generating summaries of Tatar news/articles
• Zero-shot Classification: Cross-lingual and zero-shot learning experiments
• Source Attribution: Identifying text sources

📊 Dataset Splits

The dataset is automatically stratified split into:

• Training Set: 90% (51,606 samples)
• Validation Set: 10% (5,734 samples)

The splits maintain the original class distribution.

🛠️ Data Collection

This dataset is part of the Tat2Vec project and was curated from multiple Tatar language sources spanning from 2024 to 2026:

• Merged Matbugat news articles
• Beznen articles
• Azatliq.org content
• Syuyumbike news
• Tatar-inform
• Various other Tatar language publications

📜 License

This dataset is released under the MIT License.

🤝 Citation

@dataset{tat2vec_cluster_2026,
    title = {Tatar News Clustered Dataset},
    author = {TatarNLPWorld},
    year = {2026},
    publisher = {Hugging Face},
    url = {https://huggingface.co/datasets/TatarNLPWorld/tatar-news-cluster}
}

👥 Team and Maintenance

Maintained by TatarNLPWorld as part of the Tat2Vec project.

---

Part of the Tat2Vec project - Advancing Tatar Language Processing