Chapter 9. Recommender Systems
Most online shoppers are probably familiar with Amazon's recommender system:
Figure 9.1: Amazon.com recommendations
When a customer views one item, the website displays a list of similar items that have sold well. That comes from their recommender system accessing Amazon's (amazing) database of products, customers, and sales.
Online recommender systems are now run by many vendors of goods and services: Netflix recommending movies, Apple recommending music, Audible recommending books, Yelp recommending restaurants, and so on.
A recommender system is an algorithm that predicts a customer's preferences for products based upon an analysis of that customer's previous choices compared to those of many other customers. These algorithms were pioneered by Amazon and Netflix, and are now widely used on the web.
Clustering algorithms provide one mechanism for building a recommender system: recommend what the other data points in the same cluster do. More specifically, we...
