Movie Recommender System - Notebook

Movie Recommender Systems

Recommender systems are ubiquotous in our modern world. Streaming services, social media, e-commerce, search engines, all relie on personalized recommendations for each user. Not only do good recommedations enhance user experience, they generate higher revenue for online providers. For this reason, large companies like Google, Amazon and Netflix treat their search engines as their most important company secrets and constantly work on improving the systems. To gain some basic understanding how these systems work, we use two different data sets to create a few simple recommender systems. These systems take as input a set of movies and returns similar movies.

  1. Content-Based Filtering system: We use meta data (like a general description, the director, and the actors) to find similar movies.
  2. Collaborative Filtering system: We use the review data, in particular the review scores of different reviewers, to find similar movies. We build two different collaborative-filtering systems.

Content-Based Filtering

Collaborative Filtering