We connect to a subset of DBPedia, which extracts knowledge from the infoboxes on Wikipedia articles.
So it's Wikipedia...
I'm just joking around. I love semantic databases and have worked with a few. Haven't heard of this one though, so nice link.