Skip to content

Latest commit

 

History

History
106 lines (82 loc) · 3.89 KB

README.md

File metadata and controls

106 lines (82 loc) · 3.89 KB

Semantic Book Recommender

The Semantic Book Recommender is a machine learning-powered project designed to recommend books based on their semantic relevance. It leverages Natural Language Processing (NLP) techniques to classify, analyze sentiments, and perform vector-based search for books, ensuring tailored recommendations for users.

Folder Structure

.semantic-book-recommender
│
├── .gradio                 # Gradio configuration files
├── .idea                   # IDE-specific configuration files (e.g., PyCharm)
├── data                    # Dataset files and preprocessed data
│   ├── books_cleaned.csv
│   ├── books_cleaned_with_categories.csv
│   ├── books_cleaned_with_emotions.csv
│   ├── cover-not-found.jpg
│   └── tagged_description.txt
│
├── notebooks               # Jupyter Notebooks for exploratory and model development
│   ├── eda.ipynb           # Exploratory Data Analysis
│   ├── sentiment-analysis.ipynb  # Sentiment analysis implementation
│   ├── text-classification.ipynb # Text classification implementation
│   └── vector-search.ipynb       # Vector search-based recommendation system
│
├── sbr-env                 # Virtual environment files
│
├── scripts                 # Python scripts for key functionality
│   ├── __init__.py         # Initialization for the scripts module
│   └── gradio-dashboard.py # Gradio-based user interface
│
├── src                     # Core source code for the project
│   └── __init__.py         # Initialization for the source code module
│
├── tests                   # Unit tests for the project
├── .env                    # Environment variables (excluded from version control)
├── .gitignore              # Git ignore file
├── README.md               # Project documentation
├── requirements.txt        # Python dependencies

Features

  • Data Preprocessing:

    • Processed datasets for semantic analysis, including category and emotion annotations.

  • Exploratory Data Analysis (EDA):

    • Comprehensive analysis of book data to uncover patterns and insights.

  • Sentiment Analysis:

    • Classify book descriptions and reviews based on sentiment.

  • Text Classification:

    • Categorize books into predefined genres or topics using NLP models.

  • Vector Search for Recommendations:

    • Implemented a semantic search engine for personalized book recommendations.

  • Interactive Dashboard:

    • Built with Gradio for a user-friendly interface to interact with the recommender system.

Setup

  1. Clone the repository:

    git clone https://github.com/yordanoswuletaw/semantic-book-recommender.git
cd semantic-book-recommender

  2. Install dependencies:

    pip install -r requirements.txt

  3. Launch the Gradio dashboard:

    python scripts/gradio-dashboard.py

Usage

  • Access the Gradio dashboard through the provided URL.
  • Upload book descriptions or search for books using keywords to receive recommendations.

Datasets

The data folder contains cleaned datasets used for training and testing, including:

  • books_cleaned.csv: Basic book data.
  • books_cleaned_with_categories.csv: Books with category annotations.
  • books_cleaned_with_emotions.csv: Books with emotional tags.
  • tagged_description.txt: Sample tagged descriptions for vector search.

Notebooks

The notebooks folder includes detailed workflows for:

  • Data exploration (eda.ipynb)
  • Sentiment analysis (sentiment-analysis.ipynb)
  • Text classification (text-classification.ipynb)
  • Semantic search (vector-search.ipynb)

Contribution

Contributions are welcome! Please fork the repository and submit a pull request with your improvements.

License

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