Configuring Content, Actions, and Context to make recommendations

So far in the chapter, we have given you a pretty good idea of what the Personalizer service does for its audience, and we looked at what happens in the background with the machine learning algorithms and libraries that were used to develop the service. Now, we’re going to start to dig into the service itself to look at how it is configured, and what the details that go into the environment variables we’ve mentioned a few times are by looking at Features. Note the capital “F” in that last sentence for Features. Yes, that was intentional, because when we talk about Features, we are talking about an actual concept in the Personalizer service, not just the features of the service. Get it? Sometimes, the naming decisions for these things make me laugh.

When we discuss Features, we refer to a unit or grouping of Actions or Context. Features provides all the information needed to help improve...