I designed a centrality metric to distinguish most connected nodes. The model is designed based on degree distribution of nodes. The first result implemented without edge weights show node I ranked over nodes B, D, E in figure 1 although they have the same number of connections.
The second result implemented with edge weights show that node I still ranked higher than nodes B, D, E in figure 2. I had expected that taking the weights into consideration will give more importance to nodes connected to edges of higher weights. But this is not the case.
My observation is that sometimes when nodes have the same number of connections, the model seems to penalize nodes connected to edges of higher weights and reward nodes connected to edges of lesser weights.
I would like to know if there is a natural occurrence of something like this in real world networks or is there something wrong with my model?
The correlation between two ranked list is high!
Does your model probabilistic or deterministic? I see node E as a important node, more than H, for example, B is a bridge, but for help you is important to know how your model works.
Hello Carolina Ribeiro Xavier !
Thank you very much for your input.
My model is a deterministic model. I think that the parameters I used initially in the model design are responsible for the result I got here.
I later made some modifications and I arrived at better results.
- Badges
- Science topic
- Similar topics
- Biological Science
- Bioecology
- Systems Ecology
- Landscape Ecology
- Connectivity