I have worked a bit over the past two years with social network analysis…dipping my toes in the all too alluring waters of the field. I have tried different programs (pajek, UCINET, Netdraw) and even spent a LONG month and a half trying to do things manually.
You see, the beginning of my thesis is so dependent on defining the actors and their social attributes in different bloging networks. (It will focus on sociolinguistics, with a heavy dash of corpus based exploration.) To me, what is happening with language in blogging is important…but even more important is the why and the why can only be discovered in the intersection of language and the actor. (And on a more personal note, I find SNA *fascinating*.) My main problem has been finding a way to interact with all the pretty little visualizations I was creating. I was getting pictures, but no clear way to interact with the information they represent (picture on right created with Pajek - more about what it represents can be found here).
Recently, I also received permission to look at the ICWSM dataset. I wanted to get a good look at different networks in a bigger picture. I find this opportunity exciting because so much of current weblog/CMC research has made somewhat sweeping generalities about language without looking at the language in it’s social/network context. By comparing language variation within networks and among them, one can make certain generalizations about language in weblogs, rather than language in a particular network of webloggers. This dataset is huge, and we have had to take certain measures to reduce the amount of data (looking at only English weblogs, for example, and coding for only strong ties). Once it finishes parsing today, I will run it through another compelling program I have recently found called GUESS. This java based visualization program allows you to interact with the data in the program. It seems quite intuitive and the developer very willing to answer questions. (So excited about running the parsed data I feel like it is Christmas morning!) I will post preliminary pictures here asap. It has taken me a bit of time to get used to using the terminal, but I have to admit, it was not difficult to set up at all. I can not wait to put the program through its paces!
On Nov. 17th, I will present my pilot study from this dataset to my department. If you would like to skype in, please email me (stephanie.hendrick (take out) @ (take out)engelska dot umu dot se)and let me know. I would love to have feedback on this initial paper.