PDF Document Summarizer

An automated tool that extracts text from PDF documents and generates concise summaries using the OpenAI API and LangChain.

Features

• PDF text extraction: Extract text content from PDF documents.
• AI-powered summarization: Generate concise summaries of long documents.
• User-friendly interface: Simple web interface for uploading PDFs and viewing summaries.
• Handles long documents: Efficiently processes and summarizes lengthy documents.

Project Structure

pdf-summarizer/
│
├── app.py                  # Main application file with Gradio interface
├── pdf_extractor.py        # Module for extracting text from PDFs
├── summarizer.py           # Module for summarizing text using OpenAI API
├── .env                    # Environment variables file (create from .env.template)
├── .env.template           # Template for environment variables
├── requirements.txt        # Python dependencies
└── README.md               # Project documentation

Prerequisites

• Python 3.8 or higher
• OpenAI API key

Installation

1. Clone this repository or download the source code.

2. Create and activate a virtual environment:

python -m venv pdf_summarizer_env
source pdf_summarizer_env/bin/activate  # On Windows: pdf_summarizer_env\Scripts\activate

3. Install the required packages:

pip install -r requirements.txt

4. Set up your environment variables:

cp .env.template .env

Then edit the .env file and add your OpenAI API key.

Usage

1. Start the application:

python3 app.py

2. Open the provided URL in your web browser.

3. Upload a PDF document, and click "Process PDF".

4. View the extracted text and the generated summary.

Customization

• To change the OpenAI model, modify the model_name parameter in the TextSummarizer class.
• Adjust summary length and style by modifying the prompt template in the summarize_text method.
• For better handling of specialized documents, you can extend the extraction and chunking functions.

Limitations

• The tool may not correctly process PDFs with complex layouts or heavily image-based content.
• Summarization quality depends on the OpenAI model used and the clarity of the original text.
• Processing very large documents (100+ pages) may take significant time and API tokens.

License

This project is licensed under the MIT License.