Russian News Articles With 11 Topic Labels

Over 5 years of Russian news texts are categorized into 11 distinct topics, including politics, economy, and sports. The dataset is designed for multiclass text classification tasks and is described as well-balanced. It was created by data-silence and uploaded to Hugging Face in August 2024.

Use Cases

• Train a model to classify news text into one of 11 categories like 'politics' or 'science'.
• Fine-tune a transformer model for Russian language understanding using the labeled 'economy' and 'society' articles.
• Analyze topic distribution and balance across the 'conflicts', 'health', and 'culture' labels over a 5-year period.
• Build a news filtering or recommendation system based on the 'sports', 'travel', and 'climate' topic classes.

Strengths

• Covers a 5-year time span of news content.
• Includes 11 distinct and clearly defined topic categories.
• Described by the creator as a well-balanced dataset for classification.

Limitations

• The exact number of rows and articles is unknown.
• Potential geographic bias towards news relevant to Russian-speaking regions.
• No column details or sample data are provided to verify structure.

Provenance

Source

Hugging Face user data-silence.

Collection Method

Collection of news items over a period, method unspecified.

Time Range

News from the last 5 years prior to dataset creation (exact years unknown).

Freshness

Last updated on the platform in August 2024.

Geography

Primary language is Russian, suggesting coverage of Russian-speaking regions.