Book Recommender System

Inspiration

We want to build our book recommender system for users for free and that they don't feel like they are selling out their privacy.

What it does

It will take 10 books as input and based on the backend we have trained our neural network model so that it can give back 10 books that we suggested.

How we built it

We trained our model with Python and Tensorflow.js and then build the website with HTML, Bootstrap, Jquery, and Javascript handle the client-side prediction and CSS for a 2D design

Challenges we ran into

We are all very new to web development, so developing a web application which can be deployed (from training the ML model in a python script to creating a responsive website) had a very steep learning curve. Collaborating on the same code remotely was also a challenge, as was finding a suitable data set for our needs.

Accomplishments that we're proud of

It is the first time we trained our model and made a web demo for it. In term of order training, we have managed to lower down the loss as much as possible and also with the training and validation graph, our model does not overfit by much. In term of making the website, we have learnt a lot with bootstrap and building the search engine.

What we learned

We learned A LOT about bootstrap and jQuery, as well as gaining more experience in Javascript. We also learned how to load a machine learning model into Javascript, and how we can save user input in Javascript.

What's next for Book Recommender System

The project is not done: we ran into a huge obstacle during the final steps of combining the website with the machine learning model, and to fix this it would require much more work (and it's quite late right now...). So we do intend on eventually finishing this project!