I have a bipartite graph and I want to partition each of the two sets of nodes (I don't know the technical term for these sets I mean each part of the bipartite graph) separately.

Each node hase a multidimensional location (x1, x2, x3, xn) as well I want to consider the edges on this graph in the clustering (if two nodes have the same distance in X from a third node the one that shares more connections to the nods connecting to third node is nearer to the third node).

Is there a method that is most appropriate to this situation. I have done lots of different clustering methods on points located in a continuous space but I can't see how I might apply similar methods to this.

Any comments appreciated.

Thanks in advance.


This should be helpful (article about how to partition a bipartite graph into partite sets) will help. I think you want to find the partite sets.

  • $\begingroup$ Thanks, the article you mentioned is useful. But I am not looking for "partite sets" in my graph, the graph is bipartite but I want to cluster (in the sense of cluster analysis) the nodes in each partite set into clusters based on their individual properties as well as their connectivity with members of the other partit set. $\endgroup$ – Ali Mar 12 '11 at 3:24

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.