Binge is a movie recommendation and review website built with flask as backend and implicit for generating recommendations.
This project uses ALS and approximate nearest neighbours (NMSLib) algorithm for generating recommendations.
All the data is stored in a sqlite database (app.db
). For implimentation purpose dummy users are created using moviewlens 100K dataset.
-
Recommendations based on user as well as similar movies. Write reviews and give rating.
-
Follow users to see thier reviews in feed.
-
Search movies and users (used flask-msearch)
-
Admin view (localhost:5000/admin) (only accessible by user with
role='ADMIN'
)
setup conda environment
conda create --name binge --file=environment.yml
run the server
conda activate binge
flask run
- flask
- flask-admin
- flask-login
- flask-mail
- flask-migrate
- flask-msearch
- flask-sqlalchemy
- flask-wtf
- implicit
- nmslib
- pandas
- A dummy admin is present in the database
username: susan
password: password - Other dummy accounts include
username: cooper, susan3 to susan610
password: dogs