Article Summary Generator for Telegram

A powerful Streamlit application that generates concise summaries of articles from Wikipedia and Medium based on user queries and automatically posts them to a Telegram channel. Perfect for researchers, students, and knowledge enthusiasts looking to quickly digest information from trusted sources.

Features

• Intuitive Web Interface: Built with Streamlit for seamless user interaction.
• Smart Article Search: Uses the Serper API to find relevant articles on Wikipedia and Medium.
• AI-Powered Summarization: Leverages LangChain and Ollama to create concise, high-quality summaries.
• Telegram Integration: Automatically posts summaries to a specified Telegram channel.
• Structured Workflow: Managed by LangGraph for efficient and reliable processing.

Technologies Used

Technology	Description
Python	Core programming language for the application.
Streamlit	Creates the interactive web-based user interface.
LangChain	Framework for building applications with language models.
LangChain-Ollama	Integrates Ollama for advanced summarization tasks.
LangGraph	Manages the workflow of keyword suggestion, search, and more.
Serper API	Enables web searches for relevant articles.
Telegram API	Sends summaries to a Telegram channel.

Prerequisites

• Python 3.8 or later: Ensure you have Python installed (Python Downloads).
• Ollama: A local AI server for running the "qwen2.5:latest" model. Install it from the Ollama GitHub repository.

Installation

1. Install Ollama:

   • Follow the instructions at the Ollama GitHub repository.
   • Pull the required model:

     ollama pull qwen2.5:latest

   • Start the Ollama server:

     ollama serve

2. Clone the Repository:

   git clone https://github.com/armanjscript/Article-Summary-Generator.git
   cd Article-Summary-Generator

3. Install Python Libraries:

   pip install streamlit langchain langchain-ollama langgraph python-telegram-bot python-dotenv nest_asyncio

4. Set Up Environment Variables:

   • Create a .env file in the project root with the following:

     SERPER_API_KEY=your_serper_api_key
     TELEGRAM_BOT_TOKEN=your_telegram_bot_token
     TELEGRAM_CHANNEL_ID=your_telegram_channel_id
     

   • Obtain the keys:
     • SERPER_API_KEY: Sign up at Serper API to get your API key.
     • TELEGRAM_BOT_TOKEN: Create a bot via @BotFather on Telegram.
     • TELEGRAM_CHANNEL_ID: Find your channel ID using @RawDataBot by sending /start and /id in your channel.

Running the Application

1. Ensure the Ollama server is running in the background:

   ollama serve

2. Start the Streamlit app:

   streamlit run main.py

3. Open your browser and navigate to http://localhost:8501 to access the application.

How It Works

The application uses LangGraph to orchestrate a workflow with the following steps:

1. Keyword Suggestion:
   • The user’s query is processed by the Ollama model ("qwen2.5:latest") to generate up to 3 relevant keywords.
2. Article Search:
   • The Serper API searches for articles on Wikipedia and Medium using the keywords, selecting up to 3 URLs per source.
3. Article Loading:
   • LangChain’s WebBaseLoader fetches the content of selected articles, adding metadata like source and title.
4. Summarization:
   • The Ollama model generates concise summaries (under 350 characters each) for each article, combining them into a final output with the query, keywords, and sources.
5. Telegram Posting:
   • The summary is sent to the specified Telegram channel via the Telegram Bot API.
6. Result Display:
   • The Streamlit interface shows the keywords, article URLs, summary, and a confirmation of the Telegram post.

Example Usage

1. Run the app:

   streamlit run main.py

2. In the web interface at http://localhost:8501:
   • Enter a query, e.g., "Machine learning."
   • Click "Generate Summary."
3. View the results:
   • Keywords: e.g., "machine learning, artificial intelligence, deep learning."
   • Article URLs: Links to Wikipedia and Medium articles.
   • Summary: A concise summary of the articles.
4. Check your Telegram channel for the posted summary.

Dependencies

Library	Purpose
streamlit	Web interface creation
langchain	Language model framework
langchain-ollama	Ollama integration for summarization
langgraph	Workflow management
python-telegram-bot	Telegram API interaction
python-dotenv	Environment variable management
nest_asyncio	Asynchronous operation support

Environment Variables

Variable	Description
SERPER_API_KEY	API key for Serper (Serper API).
TELEGRAM_BOT_TOKEN	Telegram bot token from @BotFather.
TELEGRAM_CHANNEL_ID	Channel ID from @RawDataBot.

Limitations

• Searches are limited to Wikipedia and Medium, with up to 3 articles per source.
• Summaries are capped at 350 characters to fit Telegram’s limits.
• Requires a running Ollama server with the "qwen2.5:latest" model.
• Ensure API keys and Telegram credentials are correctly configured to avoid errors.

Contributing

Contributions are welcome! Fork the repository, make your changes, and submit a pull request. Please ensure your code follows the project’s style and includes appropriate tests.

License

This project is licensed under the MIT License - see the LICENSE file for details.

Contact

For questions or feedback, please contact [armanjscript] at [[email protected]].