Paul Butler at Facebook recently did a really nice piece of work of displaying friendships geospatially, allowing us to see the global scale of Facebook relationships. Piggybacking off of that idea, our CEO Christopher Ahlberg thought we at Recorded Future could do something similar, and put together some visualizations looking at political relationships between countries.
Understanding shifts in the geopolitical world is not trivial. In the spirit of Paul Kedrosky's data exhaust idea - "the kinds of information we throw off as a byproduct of our actions" - here we will look at travel events of world leaders, with the theory that world leader travel is not only an indication of interest but perhaps even a leading indicator.
We used the Recorded Future News Analytics API to extract travel events on 1,000 VIPs - prominent politicians, generals, and business people from around the world - and created a data set of 350,000 events. After removing duplicate reporting, we ended up with 7,000 unique VIP trips done during 2009-2010.
Using Spotfire to visualize the trips, we can immediately see a subtle "map" emerging, very similar to the Facebook example. Washington D.C, Beijing, and a few other places stand out. The data includes both city names as well as trips starting or ending in just a country (hence the Russia node in upper right corner).
(all visualizations produced with Spotfire - http://spotfire.tibco.com/)
We can look specifically at trips originating in Beijing, essentially seeing what countries/places China really cares about.
Or a smaller country, but certainly active in its foreign relations: Iran.
But we can go a bit deeper and beyond the map. If we use a connectedness metric to filter out only the five most connected cities in the world (again assuming the leader travel is a leading indicator of connection/importance/relevance), we find that the most important cities in the world are all rising powers like China, India, and Turkey. Not only do we find strong connections between Washington D.C. and these capitals, but also pretty much between every other pair but New Delhi and Ankara. The single most powerful connection we find, unsurprisingly, is between Washington D.C. and Beijing, as befitting the new world order.
Instead of analyzing just cities we can of course look at people. Here we select Xi Jinping, the likely successor of Hu Jinatao in China, and review his connections, potentially revealing his future focus.
Finally, as an alternative visualization we can review highly connected cities and use a treemap to see which people travelled there, coloring each person by how favorably or unfavorably the visit was portrayed in the media.