API with 10,000 books (each averaging 50,000 reviews) that given a selected book, recommends books with one of two methods: Nearest Neighbors or Correlation Matrix Factorization.

prod url: books.bendevera.com

• GET /api - root route with constant response.

• GET /api/books - route to get all books in the database.

• GET /api/books/<id>- route to get an individual book by id.

• POST /api/recommend - route to get book recommendations.

  • book_id | int | [1,10000] | "reference" book that is used to generate recommendations
  • type | string | nearest/matrix/popularity | method used to generate recommendations

1. install dependencies pip install -r requirements.txt
2. create postgres database
3. set flask app export FLASK_APP=manage.py
4. initialize flask db and migrations folder flask db init
5. create first migration flask migrate -m "message"
6. upgrade db flask db upgrade
7. store all books using script in root directory python store_books.py
8. store all recommendations (preprocess for speed) python store_recommendations.py both|mf|nn
9. run app flask run