Graph Visualization For Geography - Methods for Multilevel Analysis and Visualisation of Geographical Networks - page 89

Fig. 6.7 Multiscale compound visualization of the 2000 international air traffic network. Connec-
tions with fewer than 300,000 passengers have been removed and the resulting graph clustered
using a strength clustering algorithm ( Auber et al. , 2003 ) (Image courtesy Amiel, Rozenblat, and
Melançon ( 2005 ))

London, Frankfurt) and North American airports have been clustered together
(cf. Fig. 6.7 ), because they are highly connected and form one of the main hubs
of the network. One can also see that Asian airports have been clustered together
and that these airports are sparsely connected to the rest of the network (in Fig 6.7
the “Asia” metanode is linked only to the main hub and to several other clusters).

6.4 Flow Maps Representation Of Graphs

Graph drawing algorithms focus primarily on the computation of node positions and
try to achieve various aesthetics in order to provide readable layouts. A major graph
drawing aesthetic is edge crossing minimization ( di Battista, Eades, Tamassia, &