That page for Tokyo looks suspiciously like the internals of a natural language processing system. I can't wait until someone hooks up a probabilistic speech parser to Wikidata. A talking computer with the knowledge of Wikipedia could be incredible.