Interactive visualization supports the analytical process: interacting with abstract views, using the data as a malleable material, analysts build hypothesis that can be further validated on the whole dataset. We use graph clustering in order to group elements and show them as meta-nodes and reduce the number of visual items, further organizing data over into several layers, in an effort to provide a multilevel model of the studied phenomenon. In doing so, hierarchical clustering not only contributes to the study of data but also brings in ingredients for interaction, enabling the user to zoom in and out of clusters while exploring the data in quest for evidence.