Find My Books

Semantic sentence vectorisation for Book recommendation Find the deployment here findmybooks.ml

About

A pretrained distilRoberta model used to generate vectors for description sentences of shape 768. A corpus of 50k books' descriptions are vectorised and saved as np arrays which is further used to output similar books by calculating dot product of the vectors.

For more info on the pre-trained model used:

Model name: paraphrase-distilroberta-base-v1

Sentence Embedding Model

Requirements

1. Python 3+
2. Node JS 10+

Dev dependencies

• Pandas
• Numpy
• React
• Material UI

Installation

Install Javascript node dependencies by running the following in source/js/ directory

npm install

Install Python dependencies by running the following in in source/python/ directory

pip install requirements.txt

Usage

To start the UI react app,

npm start;

To start the backend API flask server

python app.py

Contributors

Anup Kashyap TS Aditya

License

BSD

Disclaimer

Package contributors are not responsible for any misuse.