Markov-Chain based centralities and Space Syntax’ Angular Analysis: an initial overview and application

Centrality measures of Integration and Choice have performed a crucial role for Space Syntax in depicting complex relations among form, function, and movement within cities. However, while still relevant, those measures are unable to address certain innate network properties regarding the relative importance of certain road elements, essential for urban analyses focused on road-network resilience. The overreliance on Integration and Choice metrics to explain urban phenomena left several configurational patterns derived from connectivity rather unaddressed by Space Syntax and currently constitutes the methodology’s main limitation. With those points in consideration, this paper proposes an initial overview regarding the adaptation of Markov-based centrality measures to the Space Syntax framework and graph representation. These measures, often computable only in primal graphs, are adapted to the Road-Centre Line representation, in a first approach that aims to further integrate them into the Angular Analysis’ framework. We use the measures of Normalized PageRank Centrality and Normalized Kemeny-based centrality that estimate the connective relative importance of individual road-elements within the system. These measures are based on the notions of strong-ties and weak-ties, both well-known concepts in social networks; Weak-ties are important to establish bridges among interconnected communities of strong-tied individuals. In the urban configuration, weak-ties give information about crucial bridges among spaces characterized by strong-ties, areas that possess a high number of interconnected road elements – common pattern found in urban settlements. Results indicate that adapting Markov-based centralities to Space Syntax is feasible and maintains a configurational and spatial sense, hence it introduces new dimensions to be evaluated in urban-regional analysis.