A Social Network Caught in the Web
This paper by Lada A. Adamic, Orkut Buyukkokten, and Eytan Adar
covers a couple of pertinent topics in my mind:
First, it provides an interesting look at social network analysis within a university online community. In the paper, the researchers use the Club Nexus online community at Stanford to understand how students in the community interact and perceive themselves and others. The interesting part to me is that their goal is to analyze the real world based on behaviors in the online community.
Several prior studies have focused on characterizing these online interactions (Curtis, 1992; Yee, 2001), and others have attempted to measure the effect of the Internet on real life social interactions (Wellman et al., 2002a and 2002b). Our study has a somewhat different focus: While we can learn much about the online community itself, we are more interested in gleaning from it insights about the underlying real world social networks.
Second, while the focus is not on the Club Nexus online community's design, the analysis does provide an interesting look at the design of the community. A couple of things I found interesting:
Users are asked for the basic (student) information like name, email, b-day, major, year in school etc. Secondly, the users are asked to create a "Buddy List" that contributes to rapidly growing membership...
Users identified their buddies by searching for them in the Stanford directory or by entering their names manually. If a user adds a buddy who is already registered, the buddy will get a notification that the user has requested to be their buddy and can accept or decline the request. If the 'buddy' is not yet registered, they will get an invitation to join Club Nexus. This viral sign-up strategy resulted in a rapid build-up of the user base.
Further, Club Nexus asks new registrants to classify themselves by choosing 3 words that relate to their personality.
In the process of registering users were asked to select three words out of a choice of 10 to 15 describing their personalities, what they look for in friendship and romance, how they spend their free time and what kind of people they turn to for support. All users completed this section as it was required for initial registration.
I thought this was interesting because the system captures valuable information in the front end of the system- prompting members to go through the process of creating buddy lists and requiring them to choose descriptors of themselves.
I think this is an important part of building online communities- using the initial registration process to capture important information that can lead to future utilization/participation.
Initial registration is one of the few times that the system has the focused attention of the new member and this time can be used to build a profile that promotes interaction. Ryze is another great example of an effective registration process.
Lastly, the paper doesn't mention it explicitly, but Club Nexus uses weblogs and links between weblogs to enable linking between members. If you view the associated Power Point, you see a screenshot of the club where blogs are listed.
This reminds me of eCademy, where much of the online community interaction and networking comes through the use of Weblogs.
Another fascinating aspects include a peer rating system called "Karma".
A Social Network Caught in the Web
You can also view the associated Power Point presentation here.