Collaborative Filtering
Collaborative filtering is commonly used for recommender systems. This approach collects information about a user's preferences and uses that to make predictions on what they may like based on their similarity to other users who have rated similar products.
A well-known example of collaborative filtering is Amazon's "Users who bought this item also bought..." recommendation system. (source ) We use the ALS (alternating least squares) method from MLlib to compute the latent factors. (source )
You can see a visual example of how this works in the Collaborative filtering Wikipedia article.
Three operators are used for collaborative filtering:
By combining these in a workflow, you can build a customized recommender system.