Okay. Well, let’s see what happens if we split them by gender. If we split them by gender, we see that the women downloaded WhatsApp and Pokemon Go, while the men downloaded Snapchat and Pokemon Go. This tells us a bit, but not much. On the other hand, if split by occupation, we can see that the students all downloaded Pokemon Go, whereas, the people who work downloaded other apps. This is a good piece of information since from now on, whenever a student comes in, we’ll recommend them Pokemon Go. Thus, occupation is a better feature here for predicting what app will the users download. So, we can go ahead and make that decision. We’ll do it by creating a node here that says, “To everybody that goes to school, we’ll recommend Pokemon Go. And for the ones that work, let’s see.” If we forget about the students, now will look at the people who work. And now, it turns out that the gender split will help us. Because the women downloaded WhatsApp, and the men downloaded Snapchat. So, let’s do that. Let’s add another node here. The new node says “If you work, then I’ll ask for your gender. And if the gender is female, we’ll recommend WhatsApp. And if the gender is male, we’ll recommend Snapchat.” And we’re done. So quick summary. If we got a student, well recommend them Pokemon Go. If we get someone who works, we’ll ask for the gender. If it’s a woman, we’ll recommend her WhatsApp. And if it’s a man, we’ll recommend him Snapchat. Now, what we need to figure out is, how do we get the computer to measure the two features, and figure out that occupation is a better feature to split by? We will learn that later.