Detecting Bridging Ties
Posted: 14 May 2014 20:34
Dear Gephi community,
for my thesis I want to analyze user content generation behavior on Facebook ego-networks and their structural measures. Part of the small-world literature by Watts/Strogatz implies that it is a certain portion of local bridges that is optimal for the diffusion of information.
I would like to test this using the extracted .gdf files of netvizz (using Gephi or Pajek).
Let's dig a little deeper into the problem: Given the definition in the book "exploratory social network analysis with pajek" a A bridge is a line whose removal increases the number of components in the network (Noov, Mrvar and Batagelj) - But this is not really what I am looking for, as there might simply be alonger path around it.
Using the dual-circle layout plug-in and the built-in modularity (Lovain(?) community detection) I arrive at a similar visual depiction to the watts and strogatz layout, of a rewired ring-lattice.
The depiction makes clear that the desired definition should rather be "an outspanning tie" and not one that can break the network apart into its comonents".
How can I quantify these outspanning ties? I.e. I am not necessarily looking for ties that are breaking the network into a new component - there might simply be a longer path around it, but I am looking for those ties, where upon their removal the average path length would increase the most. I thought about betweenness centrality here, but where is then my "cut-off" value?
Ideal would be a measure that builds up on the community detection, quantifying the number of communities that a node reaches. Is there something like this?
Thank you for your help already!
Edit: Hey there, I saw 177 people have looked into my post already. Yet I did not get any replies
Is the problem that the question is not clear? Is the answer too obvious or too complicated? I would greatly appreciate any comment or discussion, even if it does not directly solve my problem.
for my thesis I want to analyze user content generation behavior on Facebook ego-networks and their structural measures. Part of the small-world literature by Watts/Strogatz implies that it is a certain portion of local bridges that is optimal for the diffusion of information.
I would like to test this using the extracted .gdf files of netvizz (using Gephi or Pajek).
Let's dig a little deeper into the problem: Given the definition in the book "exploratory social network analysis with pajek" a A bridge is a line whose removal increases the number of components in the network (Noov, Mrvar and Batagelj) - But this is not really what I am looking for, as there might simply be alonger path around it.
Using the dual-circle layout plug-in and the built-in modularity (Lovain(?) community detection) I arrive at a similar visual depiction to the watts and strogatz layout, of a rewired ring-lattice.
The depiction makes clear that the desired definition should rather be "an outspanning tie" and not one that can break the network apart into its comonents".
How can I quantify these outspanning ties? I.e. I am not necessarily looking for ties that are breaking the network into a new component - there might simply be a longer path around it, but I am looking for those ties, where upon their removal the average path length would increase the most. I thought about betweenness centrality here, but where is then my "cut-off" value?
Ideal would be a measure that builds up on the community detection, quantifying the number of communities that a node reaches. Is there something like this?
Thank you for your help already!
Edit: Hey there, I saw 177 people have looked into my post already. Yet I did not get any replies
Is the problem that the question is not clear? Is the answer too obvious or too complicated? I would greatly appreciate any comment or discussion, even if it does not directly solve my problem.