Tech stack: Pandas, math.
Dataset used: Movie database and User ratings dataset, download **here.**
Description: A content-based and collaborative recommendation system was developed to recommend movies to an input user according to their taste profile.
The content-based system suggests movies having the highest genre score based on the genres of movies the input user has given the highest ratings to.
The collaborative filtering system finds other users with taste profiles similar to the input user using Pearson correlation between the ratings given by input user and those by other users in the ratings database. The movies liked by users having the highest similarity are then suggested to the input user.
Resulting accuracy metrics: