Visualisation of Trading of Cereals between Countries 

This picture [1] visualizes the trading of cereals between 32 countries in the year 1994. Each country has four attributes (GDP, GDP growth, population and population growth) and is represented by a graph node. The edges stand for the trading activities, pointing from exporting countries to importing countries.

The visualization is generated by Geodesic Self-Organizing Map (GeoSOM) [2], which tries to show the attributes of the graph nodes and their connections at the same picture. Graph nodes which are closed in both attribute space and graph space are placed at adjacent area. Colors are used to denote the distances between the graph nodes in attribute space. They are linearly changing from blue, through cyan and yellow, to orange. Blue denotes the smallest variance between the weight vectors, and orange the largest. Using this method, data mapped to the same blue region are close in high-dimensional attribute space and can be considered to lie in the same cluster. From the picture, we can see that the developing countries (PER, COL, BOL) are importing large amount of cereals from the developed countries (AUS, US).

Reference

[1] Visualizing Multivariate Network Using GeoSOM and Spherical Disk Layout Yingxin Wu, Masahiro Takatsuka, Richard Webber. Poster at IEEE Symposium on Information Visualization 2006(INFOVIS 2006), Baltimore, Maryland, USA, 2006

[2] Visualizing Multivariate Network on the Surface of a Sphere. Yingxin Wu, Masahiro Takatsuka. Asia-Pacific Symposium on Information Visualization (APVIS2006), Tokyo, Japan, 2006