MovieRecommender

Table of Contents

• About the project
  • Built With
• Getting Started
• Prerequisite
• How to
• Architecture
• Contact

About the project

This is a simple movie recommendation website which built on top of stack of technologies and tools for the purpose of fun and learning.

Built With

• Python
• Amazon Corretto aka Java
• Postgres
• Redis
• Docker

Please install required python libraries that stated in requirements.txt

Getting Started

To start the project, you need to have docker installed in your machine. If you do not have it, you can download the docker at Docker

Prerequisite

• Pycharm
• Visual Studio Code
• Intellij

How to

1. Clone the repo

git clone [email protected]:KYKong1997/MovieRecommender.git

2. Build docker image

docker-compose build

3. Start docker image

docker-compose up

4. SSH into postgres container and load data into database tables

The rating data is exceeding git file size limit. You may download the data from Kaggle

docker exec -it <containerId> psql -U postgres
## copy movie.csv to container
docker cp data/movie.csv <containerId>:/movie.csv
## change to movie db
psotgres#\c movie

## Create table
CREATE TABLE movie 
(
movieId serial,
title VARCHAR(1000),
....
)

## COPY csv to movie table
COPY movie FROM 'movie.csv' DELIMITER ',' CSV HEADER

Please repeat the same process for users.csv and ratings.csv

5. Batch Prediction and save predictions to Redis

python training_model/prediction.py

6. View result from browser by going to http://localhost:3000

• You can the user recommendation by entering userId into searchbox by default is no userId

Architecture

Contact

• Kuok Yong @[email protected]