The first of these is my personal Google+ network (click on the images for larger versions):
Using the modularity statistic for automated community detection in Gephi suggests (at least) 3-4 communities in this network, with a large number of more isolated nodes and clusters.
- The largest community is my Education/EdTech Circle. This includes a distinctly separate subgroup of Irish education/edtech people who have circled me on Google+. (I didn't colour them green, Gephi did that automatically ;-)
- Next comes my Science Circle, comprising practicing scientists and science writers.
- One thing which does not show up on this graph is my University Circle, because the people in it are also mostly also in other circles such as Education and/or Science, so do not map to a clear subset. This Circle therefore is more of a push mechanism for me rather than an input filter.
- Finally, the distant upper cluster consists of (mostly) Leicester students who have crossed over from my "teaching" Google+ account, which in contrast looks like this:
This network is mostly a tightly-clustered homogenous group, as would be expected for an artificially-created community built over a short time period. I am located at the tip of the arrowhead, and my academic colleagues who converse with students via Google+ cluster nearby. The distant clusters on the upper right are other, mostly social, groups that students on the course are in contact with via Google+. Around the outside of the tight lower cluster are students who have not fully integrated into this Google+ network. We might consider them to be Visitors rather than Residents, but that would be a premature conclusion based on this information alone. (We'll be looking at that in the future). Also scattered around the periphery of the student community are a few surprises ;-) Another difference is that this account has few Public posts, whereas I frequently post publicly from my personal account, changing the distribution of followers.
Static images do not do justice to the beauty of the live, interactive Gephi versions of this data. Curating the data in Gephi feels like watching a living organism responding to the inquisition of the software parameters. The interactive data also raises the possibility of using these visualisation tools for managing student progression and engagement, as I said down at t'Guardian a few weeks ago. To give you an impression of what the data feels like, here's a video of part of the construction of the above diagram:
Massive thanks to Tony Hirst for his help with this, without whose generosity and patience this would not have been possible.