Content-Based Recommendation Engine Explanation

Content-based recommendation engines recommend results based on the metadata of the item/product.

For example, suppose a user chooses a soup with garlic and cheese. A content-based recommendation engine would recommend similar items that also have garlic and cheese. 

Advantages: Without knowing anything about the user, a content-based recommendation engine can recommend similiar items.

A straight forward appoach to recommending because it simply compares items based on how similar they are to existing items. 

Disadvantages: The first factor is that a person must start with an existing item before anything else is provided.

Content-based recommendation engines suffer from lack of exploration. In other words, content-based recommendation engines only recommend things similar to what the user already likes. For example, if it is known that a user likes a recipe that has chocolate in it then the content-based recommendation engine would recommend other recipes that also have chocolate. However, the user might also like chicken, but a content-based recommendation engine would not explore that idea.

DISCLAIMER: This is an example recommendation engine for the purposes of learning and teaching. There are many other features that could be added.