This project provides an automated news aggregator for Telegram, targeting users who want a consolidated and clustered feed from various Telegram channels. It aims to simplify news consumption by grouping similar posts and categorizing sources by trustworthiness.

How It Works

The aggregator scrapes news from specified Telegram channels, then utilizes clustering algorithms to group similar posts together. Sources are categorized to indicate their reliability, and an optional LLM integration can be enabled for enhanced features.

Quick Start & Requirements

• Install: Clone the repository, install Python requirements (pip3 install -r requirements.txt), download models (bash download_models.sh), and configure Telegram API credentials in configs/client_config.json.
• Prerequisites: Git, pip, Python 3, Docker, Docker Compose. An OpenAI API key is required for LLM features.
• Run: Start the MongoDB container (docker-compose up), run the crawler (bash crawl.sh), and then run the server (bash send.sh).

Highlighted Details

• Aggregates news from Telegram channels.
• Clusters similar posts for a unified feed.
• Categorizes sources by trustworthiness.
• Supports optional LLM integration for advanced features.

Maintenance & Community

No specific information on contributors, sponsorships, or community channels is provided in the README.

Licensing & Compatibility

The repository does not explicitly state a license.

Limitations & Caveats

The project relies heavily on Telegram channel scraping, which can be subject to Telegram's API changes or rate limits. Detailed instructions and descriptions are primarily in Russian.