Latent Clustering on Graphs with Multiple Edge Types

We study clustering on graphs with multiple edge types. Our main motivation is that similarities between objects can be measured in many different metrics, and so allowing graphs with multivariate edges significantly increases modeling power. In this context the clustering problem becomes more challenging. Each edge/metric provides only partial information about the data; recovering full information requires aggregation of all the similarity metrics. We generalize the concept of clustering in single-edge graphs to multi-edged graphs and discuss how this generates a space of clusterings. We describe a meta-clustering structure on this space and propose methods to compactly represent the metaclustering structure. Experimental results on real and synthetic data are presented.