This week’s assignment looked like it would be fun, unfortunately, after several days of struggling with Gephi and trying to get data about my Facebook network, I am having a rare moment of PC envy and extreme irritation with Facebook. I spent a few hours downloading Gephi and looking around at the sample data but since most of the video tutorials available were using Facebook networks, that could be downloaded, I decided to try to get mine. The Wolfram|Alpha link promised to get me my network but nothing was happening. Apparently, Facebook no longer allows us to access to our own network; I contacted Wolfram|Alpha and they have since taken the page down. I also tried the Netvizz App on Facebook several different ways, but wasn’t able to get anything useful. I was very disappointed because I would so like to have a picture of my Facebook network. I have over 1,000 contacts from all around the world. Believe it or not I only send friend requests to people I know but I accept them from most anyone if they have a connection to either me, my community or my other friends.
Finally, I tried some Excel data files I had downloaded and tried opening them in Gephi and it worked. I had a network file of Health Statistics and was looking at a picture of the data. I was able to change the colors of the nodes, to edit the thickness of the edges and to pull apart and label the nodes so that I could see a clearer picture. I realized that particular data wasn’t useful to look at as a network. It was death rates, birth rates and mortality and I didn’t know how to look at it differently so I started playing with some other files and foolishly didn’t think to take a screenshot. After looking at some of the sample data sets and graphs, something happened and I lost the view. I tried uninstalling and reinstalling but it keeps starting the same way. I went to the user forum and learned I need to delete the ‘user file’ but I don’t know enough to figure out where that is so I have given up for today.
Since I failed at doing anything useful with Gephi and I don’t have a PC so I can use NodeXL, I was given the option of studying the NodeXL video documentation and posting some pictures of social network visualizations. I was able to find a site that produces a network visualization of my LinkedIn network: Socialab.com and actually got a picture. The above is of the network without names, the nodes are colored by industry. The disconnected nodes are explained by the incomplete data; LinkedIn only allows 500 contacts at a time. I don’t quite understand the color coding chosen, I would say that 90% of my LinkedIn contacts are in the massage therapy industry in some way. I (labeled me) and several others are teal, yet many other massage therapists are light green and some are dark green. I am not able to do much with the visualization, I may try it in another browser tomorrow to see if there are any changes and if it is easier to manipulate.
When I first learned of Social Network Analysis as I was applying to graduate programs in 1991/1992, the technology was so limited. However, even then there was great potential to learn more about the importance of social relationships in the social sciences. To me it is so exciting to see the improvements in the technology available and I hope to learn more about the the uses of social network analysis.
Watching the tutorials for NodeXL, I found myself truly wishing I had a PC for the first time since I switched in 2009. I have been slowly learning to work with Excel for Mac and the NodeXL plugin looked as though it would be fairly straightforward to install and use, if only I could. While I haven’t been successful at getting social network data that I could play with, I still found myself wishing I could explore the uses of this program. While the little I could discover about Gephi before I messed up the GUI, seemed similar to what can be done in NodeXL. Tasks such as assigning weight and direction to the edges, filtering and separating nodes by color etc. are available but the Gephi documentation is slim and there is much more material available for learning to use NodeXL.