juliengan / Movie-Recommender

IMbD Recommender

Movie-Recommender

• Process the user ratings with PySpark
• Collaborative filtering: algorithm ALS (Alternated Least Squares)
• KMeans to recommend unseen movies to users from preference clusters
• LinearRegression to predict box office revenues and average ratings

Analysis of the ratings and movies

• Movies grouped by production names
• Movies ordered by popularity and other relevant insights

Prediction of vote_averages and box office revenues thanks to a cross validator containing a RegressionEvaluator

Recommendations of movies

• Use of ALS to recommend N top movies from its users and predict users' ratings on movies they have not seen
• Use of KMeans to recommend clustered movies to clustered users