Conclusion - Methods for Multilevel Analysis and Visualisation of Geographical Networks - page 201

Conclusion
Céline Rozenblat and Guy Melançon
This volume explored applications of networks to multilevel, multi-scale and
multidimensional aspects of spatial analysis, discussed “complex network” theories
and introduced the use of networks to represent the structural aspects of spatial
phenomena. With the aid of visual analytics tools, empirical studies of large network
samples examined hypotheses regarding the factors influencing the patterns of
interactions represented by graphs.
The set of papers included in this volume not only described new analysis and
visualisation techniques but also introduced new concepts relevant to geography and
more widely for social sciences. Many different networks composed of individual
entities were investigated: maritime networks linking ships, knowledge diffusion
networks (e.g., scientific networks linking authors), mobility networks (e.g., com-
muter and airplane passenger traffic), and control networks (e.g., multinational
firms' ownership of subsidiaries). In the studies presented in these chapters,
networks were aggregated based on territorial attributes to group individual entities
into new entities. In these new networks, the interacting entities were places or
territorial units, and the individual entities of the original network generated the
attributes of the newly formed entities. For example, aggregating companies based
on their geographic location produced a set of cities that inherited attributes from
the companies located in those cities (Chap. 10 ).
Using the processes designed to represent society and individuals to represent
space has many implications for the analysis and representation of the results. This
representation influences the meaning of the geographical interactions observed