Recent Developments in Complex Network Analysis in Spatial Planning

In the last years, we acknowledge a great scientific interest on complex network analysis, a method able to char-acterize systems with very large number of entities (the nodes or vertices) interlaced by a series of connec-tions/relationships (the links or edges). The objects of analyses as such are biological (predator-pray); informa-tion (Internet); social (actor-in the same movie); and transportation (railway and road networks) systems. While in general a network is an abstract- (topo)logical object, spatial networks belong to an important class of sys-tems that include nodes and edges with a clear reference to space. Recently the interest of scientists has focussed on methods able to define and investigate on communities emerging from the structure of a network. In this re-spect the spatial factor can emerge both as the result of the topological community structure that maps back onto geography in the form of sensible spatial regions, or just as spatial clusterization of nodes in principle embedded in space. In this essay, the authors aim at presenting a state of the art summary of the last advances in the field of net-work analysis and network community detection methodologies with a detailed view to the case of spatial net-works. Secondly, the paper will report on a case study concerning a major issue for policy makers and planners: the delimitation of sub-regional domains showing a sufficient level of homogeneity with respect to some spe-cific territorial features. We compare some intermediate administrative bodies of the island of Sardinia (Italy) with the patterns of the communities of workers and students, by applying grouping methodologies based on the characterization of the Sardinian commuters’ system as a complex weighted network. This essay unfolds as follows. In the next section, we develop on a brief state of the art summary on social networks with a focus on Research and Development (R&D) networks. At the end of this section, we introduce the reader to main concept of the essay, i.e. spatial networks displaying a clear geographical reference. In the third section, we review the recent advancements in the field of complex network analysis as well as its adoption in geography, spatial and regional planning. In the fourth section, we report on the latest advances regarding community detection methodologies able to cluster nodes into homogeneous groups. The fifth section presents a case study about the application of network community detection approach to study the problem of regionalisa-tion. Commuter basins in the island of Sardinia (Italy) are used to scrutinise the relevance of administrative sub-divisions at the provincial level.