This won "Best Visualization" at Matt Patterson's History Hackday in January, 2011. To make it, we parsed an xml dump of all wikipedia articles (30Gb) and pulled out 424,000 articles with coordinates and 35,000 references to events. Cross referencing these produced 15,500 events with locations. Then we mapped them over time.
Dataset: Google Fusion Tables

The dataset has 424171 events attached.