I'm studying large scale networks and graphs and I need an application for detecting dens sub-graphs. I find MCODE software and run it for some data. Can somebody introduce any other application or algorithm to me?
How large graphs are we talking about? Are the graphs fully connected? If the parent graph is fully (or at least closely ) connected I would suggest making a rectangle (or a circle) and sweeping it through the entire Parent graph. At 1 timestep / iteration you count the number of nodes that fall within your circle. Then, count the edges of each node and re-initialize the circle's center around the most densely connected node. By "dens subgraph" you actually mean "densely-connected", right? Please, let me know what you think.
We have a publication about discovering dense subgraphs: http://arxiv.org/abs/1208.1454 - I don't know if this will directly help you but the references may be useful for you.
Thank you Amitabh and Steve
Steve: graphs are sparse. a "dens subgraph" is a subgraph like or near a clique but it is not a clique and we can make a clique from it by adding some edges.
- Badges
- Science topic
- Similar topics
- Computer Science and Engineering
- Algorithms