Skip to content

Latest commit

 

History

History
95 lines (60 loc) · 2.14 KB

README.md

File metadata and controls

95 lines (60 loc) · 2.14 KB
Raw

README

Project Overview

This project provides a streamlined method for summarizing scientific articles using advanced natural language processing capabilities. By leveraging Google Generative AI and fine-tuned prompt engineering, the application generates structured summaries of uploaded scientific articles.

Features

  • Automatic Summarization: Converts uploaded articles into structured summaries.
  • Detailed Outputs: Includes key findings, objectives, methods, results, conclusions, and key concepts.
  • User-Friendly Input: Users can upload files for seamless summarization.
  • Powered by Generative AI: Utilizes Google's Gemini AI model for content generation.

Installation

Prerequisites

Ensure you have the following installed:

  • Python 3.8 or higher
  • pip (Python package manager)

Dependencies

Install the required libraries using:

pip install streamlit python-dotenv google-generativeai PyPDF2

Usage

Setting Up API Key

  1. Create a .env file in the project directory.
  2. Add your Google Generative AI API key to the file: 
    API_KEY=your_api_key_here

Running the Application

  1. Save the code in a Python script file (e.g., app.py).
  2. Run the script: 
    python app.py

Workflow

  1. Upload a scientific article file (PDF or text format).
  2. The program will process the file and generate a structured summary.

Example Prompt

The summarization prompt used in the application ensures a consistent structure:

# Structured Summary

# Key Findings

# Objectives

# Methods

# Results

# Conclusions

# Key Concepts

Code Explanation

Environment Variables

The .env file is used to securely load the API key.

Functions

  • summarize(file): Generates the summary using the Google Generative AI model.
  • main(): Entry point of the application.

Contributing

Contributions are welcome. Please fork the repository and submit a pull request.

License

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

To-do

  • translate text - ptbr
  • add copy button
  • add tutorial
  • add loading bar