Blog Tribe Social Network Mapping Here is the initial social network analysis of the Blog Tribe at Ryze -- which maps the Friendship networks and Blogrolls of Tribe members. What's unique about this collaborative project is the mapping of blogspace and of how two unique communities intersect.
Thanks to the contributions of Valdis Krebs, Pete Kaminski, and those who volunteered to contribute their blogrolls, this is one of the largest online communities mapped. The data was captured within a month of the founding of the Blog Tribe, a snapshot in time that will be useful to return to, with 1,108 nodes from approximately 100 members.
|
Ryze is an online community with explicit identities for the primary purpose of business networking.
One of the key features is the ability to add "Friends," links from your identity page to another. Pete developed a web mining script to capture the links from each member's page to their Friends. The results captures the membership's network in its entirety.
Using Orgnet's InFlow 3.0, a social network mapping and measurement tool, Valdis generated the image to the right. |
 |
 |
Most members of the Blog Tribe maintain weblogs. Members were requested to volunteer a list of of weblogs they link to (commonly known as a Blogroll). Since the formats of weblogs are not standards and tracking down the names behind the weblog is cumbersome, we were not able to automate this task. Given the time cost, only 16 members made an initial data contribution, which I manually collected.
The image to the left, generated by Orgnet, may be the first social network map of a weblog community by name of blogger. |
Comparing the differences between the fitness of a weblog and a Ryze page node may allow us to infer how these two distinct networks will evolve. Unlike Ryze, Weblogs are not currently designed as networking tools, as: Identities are not explicit (save for a person's URL), and transaction costs are higher for linking from an identity to another. Aside from networking, the absence of these characteristics is a strength in and of itself. Non-explicit identities appeal to many users and content-to-content linkages are formed easily.
Valdis goes further to provide us views of Network Horizons:
|
Network Horizon – how many nodes are within X steps of you?
X is usually 2 or 3 – two being a tighter definition. The more nodes, within 1, 2, or 3 steps of you, the larger your network horizon. Here size does matter, but it is not everything! Diversity and heterogeneity of the nodes within your horizon also matters. After all, combining many of the same items may give you increased mass, but it won’t give you anything new and interesting. Dipping into complexity science… the more recombinations, with more diverse elements, the greater possiblity for variety, innovation, and opportunities. |
The following tables demonstrate the difference between 2 and 3 steps:
|
Network Horizon of 2 Steps
[Score = % of all nodes]
0.832 Ross Mayfield
0.668 Adrian Scott
0.433 Ken Berger
0.422 Miko Matsumura
0.420 Mayuko
see full list... |
|
Network Horizon of 3 Steps
[Score = % of all nodes]
0.991 Andrius Kulikauskas
0.981 Rick Bruner
0.980 Jon Lebkowsky
0.980 Ross Mayfield
0.977 Jacob Shwirtz
see full list... |
Next steps in this project include inter-network visualization, further automation of data collection, and when the Friends and Blogroll data are comparable to reveal the interrelationships between these two distinct but similar link-rich online communities.
Feedback is appreciated. How do you think these networks will evolve? What is being revealed? What would be the most useful thing for you to see? How would you contribute?
9:24:49 AM 
|
|
