---
dataset_info:
  features:
  - name: id
    dtype: string
  - name: author
    dtype: string
  - name: year
    dtype: int64
  - name: title
    dtype: string
  - name: abstract
    dtype: string
  - name: article
    dtype: string
  - name: sum
    dtype: string
  splits:
  - name: train
    num_bytes: 37252397
    num_examples: 980
  download_size: 18749417
  dataset_size: 37252397
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
task_categories:
- summarization
language:
- ru
pretty_name: humarticles
---
# **Humarticles: Russian Academic Articles Dataset**

**Humarticles** is a comprehensive collection of Russian-language academic articles, designed to support the development of AI models for text summarization, semantic search, and natural language understanding. This dataset contains a variety of academic papers from different domains, complete with abstracts, full articles, and metadata such as author names and publication years.

### **Dataset Overview**

- **Language**: Russian
- **Size**: 980 academic articles
- **Fields**:
  - **ID**: A unique identifier for each article.
  - **Author**: The author(s) of the article.
  - **Year**: The year the article was published.
  - **Title**: The title of the article.
  - **Abstract**: A brief summary of the article.
  - **Article**: The full text of the article.
  - **Summary**: A more concise summary or key insights from the article.

### **Purpose**

Humarticles was created with the goal of advancing research in several areas, including:

- **Text Summarization**: Use the articles and their summaries to train models that can automatically condense scientific papers.
- **Semantic Search**: Enhance search engines to help researchers find relevant articles quickly using natural language queries.
- **Machine Translation**: Improve machine translation systems by providing high-quality, domain-specific Russian text and corresponding summaries.
- **Linguistic Analysis**: Analyze the style, tone, and structure of Russian academic writing.


### **Contributing**

We welcome contributions! If you have any suggestions or improvements for this dataset, feel free to open an issue or submit a pull request.