Social networks that matter: Twitter under the microscope

Social networks that matter: Twitter under the microscope
Bernardo A. Huberman, Daniel M. Romero, and Fang Wu

First, I'd like to point that this article comes from a free online peer-reviewed journal called First Monday. There's a lot of internet research on this journal - a lot of it is pretty good. Anywho, I came across this article while watching a lecture from Jon Kleinberg where he discussed saturation curves in human behavior online (and perhaps elsewhere). Saturation curves like this intrigue me and you can find this phenomenon in number of places.



What we see here is that you post more on Twitter if you have more followers, that is, until about 300 followers. Then you reach saturation and there is suddenly a lot of variation in the number of posts (although, that might just be from a smaller sample size at those scales, there are a lot fewer people with 500 followers than there are with 20 followers). The point of the article was that the type of links on Twitter are important. There are two different types of links between people on Twitter, friendships and followship - friendship establishes you as a friend with a higher degree of sharing, while following just means you're listening to the tweets of the person followed.



There is no saturation point with respect the number of friends. Of course, there might be a saturation point, but there aren't very many people with more than 300 friends. We can say across the empirical range of numbers of friends, there is no saturation point with respect to number of tweets. So the authors suggest that some networks based on some kinds of links are more important to people than other networks. This is an unsurprising result, but the way that it was discovered was neat. I'm in love with saturation curves, did I mention that?