Geography Reference
In-Depth Information
Fig. 8.2 Stroke-based urban
road network
8.2.3
Community Based Dual Graph
Recently, some researchers found that the group of highly related nodes can provide
more structural information beyond single nodes (Fortunato 2010 ). Although the
stroke-based road network combines the topological related road segments into a
group, the combining rules make the group less correlated for the end nodes if the
street contains a long series of segments. Here we proposed a community based dual
graph to represent the groups of highly connected road segments.
Communities, also called clusters or modules, in complex network theory refers
to a densely connected subset of nodes that is only sparsely linked to the remaining
network (Gulbahce and Lehmann 2008 ; Fortunato 2010 ). The procedure to group
these highly related nodes in the network is called community detection. By network
clustering, nodes with many connections are grouped from original networks. In this
way, one attains a coarse-grained description of the original network from a modular
view (Fortunato 2010 ). Communities in social networks such as families, friendship
circles, and scientific collaborations have been studied for a long time (Moody and
White 2003 ).
However, little attention has been paid in community structure in urban road
networks yet. Actually, city road network also contains modules like traffic zones
or business districts. Road segments in these areas are densely connected with each
other. Communities provide a novel way to gather topologically densely connected
road segments. In this paper, we use a label propagation based community detection
algorithm proposed by Raghavan et al. ( 2007 ) to identify the communities in city
road networks and get more insights into the structural properties of road networks.
The generating procedure of community-based dual graph is shown in Fig. 8.3 .
Segments clustered in the same community are rendered with the same color in
Fig. 8.3 a. Figure 8.3 b is the corresponding dual graph. Communities in Fig. 8.3
Search WWH ::




Custom Search