So I finally completed the Social Network Analysis course by Lada Adamic today and I learnt quite a few things Alhamdulilah. Some of the MOOCs are really good but there are so many options that I sometimes get overwhelmed.
One of the cool things about the course is that students can get exports of their entire Facebook network (export tool available at this link). Of course, I got mine – I have always wanted to and I am fascinated by social computing. Analyzing my network in Gephi led to quite a few unexpected discoveries. Now, don’t get any ideas, I ain’t talking about those here ;). I’ll only talk about the ‘safe’ stuff which is already obvious public anyway.
So, I have about 1051 friends who share 23915 edges who can be segregated into four communities – these communities kind of map the friends I have made in different places and times. The average path length is about 3 while the network diameter is 8 (quite close to the much-touted 6 from the classical work of Milgram). The average degree is about 45, so my friends have on average about 45 friends? I dunno…
I generated a visualization of the network using the Fruchterman-Reingold algorithm (it provided the ‘nicest’ picture); the colours indicate the degree (i.e. number of connections to other friends) while the sizes of the nodes shows the betweenness centrality. This betweenness centrality can be explained as a measure of how important a friend is in bridging/connecting separate parties or groups of people.
The cluster on the top left shows the friends from my high school (Federal Government College, Kano). The person with the largest betweenness value (the big yellow circle) is the only person I spent about 10 years of my life with in the same institutions! (High School and University) It is quite unsurprising the person connects these two communities together.
On the top right is the community of friends I met at the MASDAR institute while the lower right (that appears to be two really close communities) are the friends I made in university – Obafemi Awolowo University. The left half is something of a sub community – it consists majorly of my colleagues in the Computer Science and Engineering Department while nearly every other person falls in the right half. Lastly, my family appears on the network too but I ain’t mentioning :).
Twitter next? Maybe.. Maybe not.