The search itself does not look terribly exciting (search for Esperanto, the left category is a bit confusing without context).
But a making-of blog article explaining the data structure and import process would be quite interesting. Is this backed by Solr (the lists could be facets, I guess)?
You really just need a wikipedia dump, then construct a graph, wouldn't be too hard. I am a bit curious how it's stored in memory and what technology it uses though.
Where are the external links? It just keep me searching. PS. The footer will overlap a long list.
It's a visual representation of Wikipedia article categories.