MovieFreak
https://moviefreak-flask-new.herokuapp.com/
installation
install postgresql for your machine
start the service using
service postgresql start
sudo -u postgres psql
then create a database using
CREATE DATABASE ratings;
then import the sql file
sudo -u postgres -i psql ratings < ratings.sql
run python app.py
IMPORTANT CSV AND JSON FILES
Some of the csv files are previously created to give recommendations using communities.
- user-com.csv - In this csv file we store the community id of each user id which was previosly calculated using Neo4j by applying Louvain algorithm on the User graph.
- small-movies-similarity.csv - In this csv file we have movie id and in the sim_movieId there is the id of the movie which it is similar to calculated based on tags, genre and rating relevance.
- community-movies.json - In this json file the key is the community id and value is a string of movie ids that have been watched by the users belonging to that particular community.The movie ids in the string are separated by commas.
- ratings.csv - It contains the information of user's rating to a movie and the timestamp of the rating.
- users.csv - UserID of all the users in the data set.
- users_genres.csv - UserID and their favourite genre . The favourite genre is the most watched genre found using movies-genres.csv file.
- users_movies - Extension of ratings.csv . Contains userId and movie they have watched to create a graph between user and movies.
- user-user.csv - The columns are user1 and user2 . Each row represents an edge between the users . This was calculated by calculating the Pearson Correlation Coefficient of the users who have watched atleast 10 movies common . The edge is only formed if PCC is greater than 0.75.
- movies_genres.csv - Relation between movies and their genre.