A system that predicts a rating user might give to an item.

Recommender system can recommend products, content, music, etc.

Two types of ratings

  1. Explicit - ratings provided by users themselves (like star rating in Amazon). Problems with explicit ratings:
    • It requires extra work from a user
    • different people have different standards, so the same rating can have different meaning
  2. Implicit - user actions can be indication of interest in specific product (like purchasing the product or opening page with product details)

Types of recommender systems:

  1. Top-N recommender

While working on the recommender system it’s important to measure it’s accuracy. Evaluating recommender system