Book Recommendation App Using Python TF-Keras and Javascript Tensorflow.js

The Book Recommender application is built using Tensorflow-Keras on book data from Here.

Link to live site here

The model uses embeddings to learn the mapping of books and users to ratings. This can help us identify similar books depending on the book and user interaction.

The Python-Model folder contains all Python training files, including dataset, saved model, and embeddings which we visualized below, using the Tensorflow Projector.

The recommender-books folder contains the web application. We first converted the saved model to Javascript format, and then embed it in the application.

To run the application, follow the steps below:

• Ensure you have Nodejs installed. If not follow the steps here
• Once you have node, clone the repo:

    git clone https://github.com/risenW/Tensorflowjs_Projects
    cd recommender-sys

• Change directory into the web app folder

    cd recommender-books

• Run the app:

    npm start

The rec-book-firebase folder contains the modified version of recommender-sys which is hosted on Google Firebase.