When we tend to think about social networks we tend to think online, things like Facebook, Twitter, and LinkedIn. And while these social media technologies have made it easier for us to see our social networks, social networks have been around for much longer then just the internet. We have friendship networks, networks of people who know each other. There are phone networks, networks of people who have called each other or a network of people who have worked together. Office mates for example over time. One interesting question is in these networks, is everyone connected to everyone? Is it possible to reach anyone from any other person in the network? And if so, how long would it take? When we think about these social networks, is everyone connected to everyone? Are we all part of one big social graph that going through friend into friends to friends of friends you could eventually be connected to anyone in the world? Is it possible to reach any one person from any other one person in any country? And if so, how long would it take? You might think it's impossible. You personally might not know anyone who lives in India or South Africa or New Zealand, but maybe if your friend knows someone it allows that connection to be made. Or it might take two or three hops Well let's start to dive into the science and understand how to answer these questions. Think for a moment about your own network. Imagine I ask you to draw a picture of everyone you know and all the people they know. Would there be separate islands of friends? What denotes those islands, and why aren't they more connected? How dense are the ties, whether within a group of people, within a given island, or between groups of people between different islands? To help us think about this, I wanna introduce you to a game called the Oracle of Bacon, or sometimes known as Six Degrees of Kevin Bacon. What it does is connect actors based on whether or not they've appeared in the same movie together. So take Kevin Costner, to help us think about this, I wanna introduce you to a game called the Six Degrees of Kevin Bacon. Kevin Bacon is a relatively famous US actor and this game connects actors to Kevin Bacon based on whether or not they've appeared in the same movie together. Take Kevin Costner for example, you might know him from Dances with Wolves fame. Well, he's in JFK the movie with Kevin Bacon so they're connected only one degree of connection between them. They are directly connected. They have appeared in the same movie together. But pick another actor, or actress that you might imagine as further away. Take Marilyn Monroe, for example, well she was alive way before Kevin Bacon was and she was never in a movie with Bacon, but she was in Don't Bother to Knock in 1952 with Eda Reiss Marin who was in a different movie called Enormous Changes at the Last Minute in 1983 with who other than Kevin Bacon. So, we're starting to see that many people can be connect to Kevin Bacon pretty quickly. Why might some people be closer or further away from Kevin Bacon? Well, you might begin to think about what makes certain networks tighter, versus looser. They're the genre of movie. If they're in different genres they might be further away. There's different time periods. There's different geographies. You might think someone like Charlie Chaplin for example, because he was alive so long ago and was in a different genre of movies, might be much further away. Well he was in a movie called Countess from Hong Kong in 1967, with Tippi Hedron, who was in Jayne Mansfield's car in 2012, with none other than Kevin Bacon. In fact to get more degrees of separation acquires a bit of work. There's an old French actress and director who did movies early in the 1900s called Mosidora. They were in a number of movies. One was called Les Vampires in 1916 with Edmund Breon who was in At Sword's Point in 1952 with George Petrie who was in Planes, Trains and Automobiles in 1987 with Kevin Bacon took three degrees of separation to get there. Now you might say, well what does movies have to do with social networks? It's fun, but what does it tell us? Well, it points out a few things. First of all, just like social networks, in this movie network, not everyone is just one hop away. Sometimes it takes multiple degrees of separation to connect people into our networks. But we can get to many people in just a couple of hops. And further, people tend to be grouped with similar others, whether they're from the same country or the same time period or other cases in movies. The same with our own lives. We tend to be friends with people that are near by where we've lived before. Same city, same country, same college, or other place that we spent a lot of time. And also, we don't always know the shortest path in these social networks. You couldn't probably have guessed how many steps it would've taken to connect Kevin Bacon and Marilyn Monroe, and just like that you might not know how many people you're connected to in a social network. The same is true for your own network. There are multiple islands if you can think of them. They might be your work colleagues, it might be your college friends, and it might be the town you grew up in In each of these islands, the ties between people are dense. Many people in your hometown are probably friends with one another. But the ties between these islands are probably much weaker. There are not many people that both went to your hometown and went to the same college as you, unless maybe you went to college very close to your hometown. The closer those two areas are to one and other the more likely there will be to be ties between those different islands. And not everyone knows every one else. Why? Well there is limited time and distance. We don't have enough time to get to know everybody. We only get to know a small set of the people that are around us. But importantly what all this points out is that ties aren't random. People tend to be friends with others like them, it is called homophily. We tend to be friends with others that are similar to us. And there are two key reasons. First, biased interaction. If you play a lot of soccer, you tend to hang out with people who play soccer, and you're more likely to become friends. If you like going to movies, you tend to go to movies with other people who like movies, and you tend to become friends. But among those people that are similar to you, who you engage in similar activities with, you tend to form ties with others that are like you to begin with. And so further and further we come more connected to people that are similar to us. So as we've been discussing, to define what a network is, it's composed of two things. Nodes, people in a case of a social network, or actors in the case of a movie network, and connections, sets of lines or edges joining those different nodes. It can be friendship, people that are friends with one another in a social network Or in the case of a movie, people that acted in the same movie. So far in this course we've talked about things that people share. And the ideas that more viral items spread further. In a social network, if people are connected, things that are more transmissible, more contagious, will spread further beyond the set of social ties they start to a larger set of social ties but the pattern of ties are also important. If there's not a bridge between two areas in a network, it's unlikely that information is gonna spread. So if we take identical items, the structure can determine the success. Certain network structures might be more likely to spread information and others might not
