So far, we have discussed content-based and collaborative-filtering-based recommendation engines. Both types of recommendation engines can be combined to create a hybrid recommendation engine. To do so, we follow these steps:

• Generate a similarity matrix of the items.
• Generate preference matrices of the users.
• Generate recommendations.

Let's look into these steps one by one.