Skip to content

Latest commit

 

History

History
143 lines (120 loc) · 5.41 KB

README.md

File metadata and controls

143 lines (120 loc) · 5.41 KB

🔖📖 BookFlow: Multi-Agent based Book Recommendation System

A hybrid recommendation system that combines collaborative filtering, content-based filtering, and sentiment analysis with reinforcement learning to provide personalized book recommendations. This system uses real-time feedback and continuous learning to adapt to users’ preferences dynamically.

Table of Contents

Project Description

This book recommendation system is designed to offer users personalized book suggestions based on their interaction history, demographic data, and review sentiments. The system uses reinforcement learning to select the optimal recommendation method (collaborative, content-based, or sentiment-based) dynamically, improving accuracy and user satisfaction.

Key Features

  • Hybrid Model: Combines collaborative, content-based, and sentiment-based filtering techniques.
  • Real-Time Adaptability: Reinforcement learning optimizes recommendations dynamically.
  • User Sentiment Analysis: Analyzes user reviews to enhance personalization.
  • Scalable Database Integration: MySQL database stores user, book, and interaction data.

Architecture Overview

  1. User Interface Layer: Manages user interactions (registration, login, book search, and reviews).
  2. Recommendation Engine Layer: Combines the collaborative, content-based, and sentiment-based filtering agents.
  3. Data Management Layer: Uses MySQL to store user data, ratings, reviews, and book metadata.

Refer to the architecture_diagram.png file for a high-level overview of data flow.

Technologies Used

  • Python: Core programming language
  • Flask: Web framework for developing the application
  • SQLAlchemy: ORM for managing database interactions
  • MySQL: Database management system
  • PyTorch: Used for reinforcement learning
  • Scikit-Learn: For collaborative and content-based filtering
  • Hugging Face Transformers: Sentiment analysis with BERT model
  • Pandas and Numpy: Data manipulation and processing

Installation

Prerequisites

  • Python 3.8 or higher
  • MySQL server
  • Git

Setup Steps

  1. Clone the Repository:

    git clone https://github.com/your-username/book-recommendation-system.git
cd book-recommendation-system

  2. Set Up Virtual Environment:

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

  3. Install Required Packages:

    pip install -r requirements.txt

  4. Set Up MySQL Database:

    • Create a MySQL database: 
      CREATE DATABASE book_recommendation_db;
    • Update your configuration file (config.py) with the database credentials: 
      SQLALCHEMY_DATABASE_URI = 'mysql+mysqlconnector://root@localhost/book_recommendation_db'
SECRET_KEY = 'your_secret_key'

  5. Initialize Database Tables:

    flask db init
flask db migrate -m "Initial migration."
flask db upgrade

  6. Download and Set Up the Dataset:

    • Download the datasets for books, users, reviews, and ratings.
    • Place them in the data/ folder.
    • Run any necessary data preprocessing scripts if needed.

Configuration

Create a .env file with your environment variables:

FLASK_APP=app
FLASK_ENV=development
DATABASE_URL=mysql+mysqlconnector://root@localhost/book_recommendation_db
SECRET_KEY=your_secret_key

Usage

  1. Run the Flask Application:

    flask run

  2. Access the Application: Open a web browser and go to http://127.0.0.1:5000.

  3. User Actions:

    • Register and Log In: Users can register and log in to the application.
    • Search and Review Books: Search for books by title, author, or ISBN, and submit reviews.
    • View Recommendations: See personalized book recommendations based on user preferences and feedback.

Project Structure

book-recommendation-system/
│
├── app/                       # Flask application files
│   ├── __init__.py            # App initialization
│   ├── models.py              # Database models for Users, Books, Ratings, Reviews
│   ├── routes.py              # Application routes
│   └── templates/             # HTML templates for UI
│
├── data/                      # Dataset files
│
├── agents/                    # Recommendation agents
│   ├── collaborative_agent.py # Collaborative filtering logic
│   ├── content_based_agent.py # Content-based filtering logic
│   └── sentiment_based_agent.py # Sentiment analysis logic
│
├── rl_model/                  # Reinforcement learning components
│   ├── actor_critic.py        # Actor-Critic model for agent selection
│
├── config.py                  # Configuration settings
├── requirements.txt           # Required packages
└── README.md                  # Project documentation

License

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

⚠️ Note: Make sure to replace placeholder text, such as your-username and your_secret_key, with your actual information.