Click "Show" in the applet below to show Dagstuhl-MLSW as a graph, where nodes are abstracts in the proceedings, and there is an edge between two abstracts if they are in the same area/theme (Information Extraction, Ontology Learning, etc).
To play with the automatic layout algorithms, click the blue button in the lower-right corner of the graph window.
Here's a reasonably informative static visualization:
Thanks to RPI for their graph-layout code. If you're curious, here is the raw data.