Linked by Albert-Laszlo Barabasi
Networks
Linked: How Everything Is Connected to Everything Else and What It Means, by Albert-Laszlo Barabasi, is all about networks, which is what we were all about last week at the UX Book Club in New York. As people were networking with each other, they were asking interesting, difficult, and sometimes unanswerable questions, in order to make sense of what often is an intangible, tangled web of nodes and links.
One of the principal topics we explored was whether the theory of networks could be extrapolated to other domains, the exercise put forth was to imagine that we have been charged with the task of designing a network for a company. Although quite vague, it gave way for very diverse types of discussions.
Creating vs Finding
Someone started with the excellent insight that networks already exist; therefore, we don’t need to re-create them. Our task is simply to make a tool that fits the network. The main question now becomes, how do we find this existing network? Before answering this, we must take a step back and identify our variables. For the context, assume we are talking about people, as in the ever popular social network. Our purpose is to broadcast information to the greatest number of people with the least amount of resources. Clearly then, the first step is to find the hubs, those few people with a disproportionally greater amount of links than the majority. By identifying a handful of highly connected people, the size of your network suddenly grows much faster and much larger than if you identified a lot of random, averagely connected people.
Hubs
But how do you find the hubs? Two analogies formed the basis of one idea. In PET scans, a radioactive particle is attached to a protein of interest, then sent through your bloodstream, where it traces the path that this protein travels. In programming, there is practice called tracer bullets, where cascading output is sent through different parts of the program to trace its path and discover the root of an error.
Thus, we send a tracer particle/bullet through the network, gradually revealing an ever-growing snapshot of the network as it traces its path through it, and eventually leading us to the hubs we are seeking.
Hub Hubs
Then someone else made a great observation. You don’t want to target just hubs, you want to target hubs who are more closely linked to other hubs. A hub linked to other nodes who have no links themselves is not as useful as a hub, maybe a smaller one, that is directly connected to more hubs. In other words, those with the minimum distance to the maximum number of hubs, or hubs of hubs.
Concrete examples
Pretend your company just came out with a new app. You implement a referral program, where a person receives a reward (a point, a credit, or a dollar) if they refer someone to your app. By keeping track of who refers who you begin to light up the network, node by node, link by link. Finally, you discover a hub, a person who referred ten times more people than anyone else. Thus, you invest more resources there and target the hubs specifically, such as offering them more incentives or greater rewards.
Another theme, which kept re-surfacing throughout the conversation, was Twitter, and specifically, re-tweeting. A solid strategy would be to track down those people who have the most followers and re-tweet your tweets the most. But it doesn’t end there. You want to find those people whose re-tweets get re-tweeted, that is, those that generate the greatest amount of 2nd and 3rd degree re-tweets.
In Sum
You don’t need to be an expert mathematician to understand that networks are important, nor to be able to use them to your advantage in whatever you are doing. If you are looking for an introduction to networks, and insights into their prevalence in our world, read the book Linked.
In an attempt to put theory into practice, leave a comment with your name and how you got here. Let’s get Linked…
Tags: networks
3 Responses
Leave a Reply
Blogroll
- Next >> Back to Blogging
- << Previous "Personas"
It is what we encounter every day. Following on your hub and hub of hubs point, when looking for these hubs of hubs you also need to be careful. What if you encounter two hubs of hubs whose all their links are in common? Then you only have one hub of hubs. Imagine, you have the referral program and you find a hub of hubs with “x” connections, this hub refers “x” people, but these people have a good number of connections in common and connections start overlapping with each other. Thus, when you are looking for hubs of hubs you need to build a strategy to filter out these overlapping ones which will not lead you very far.
I’m currently an MBA student. The importance of networking and networks have been drilled into our skulls. In business we not only identify the hubs but also the “super” hub. Once you have the super hubs make sure you keep them. Interesting article, I will have to look into the book.
Thank you very much for your help, this site has been a great relief from the books,