You might be interested in this http://dbpedia.org/docs/ – it is an attempt to automatically extract structured data from Wikipedia, and present it in a computer-friendly form.

It is then possible to download the dataset, or search using a SPARQL (semantic web/RDF query language as far as I know) query:

eg. http://wikipedia.aksw.org/index.php?qid=13 (there are more down the left-hand side)

They appear to have accumulated a rather large dataset: http://dbpedia.org/docs/#down