| Summary: | Using conventional relational data (residential migrations, commutes to and from the workplace) and less conventional relational data (mobile telephony calls), the space in and around the Brussels-Capital Region is partitioned into groups of closely inter-related places using a mathematical community detection method. The partitions obtained lead to strong spatial structures, while neither the distance nor the characteristics of the places are taken into account in this method. This article illustrates how large databases (big data) and their specific methods provide new opportunities for urban analyses (delimitation of urban borders, formalisation of intra-urban structures), and remind us here that no structure may be interpreted without a thorough understanding of data, the tools used and regional and urban theories.
|
|---|
