RANKING SPACES FOR PREDICTING HUMAN MOVEMENT IN AN URBAN ENVIRONMENT
Department of Land Surveying and Geo-informatics, The
In this paper, we consider a city as an interconnected graph or topology in which nodes and links represent individual spaces and space intersections respectively. Based on the topology, we apply weighted PageRank algorithm for ranking the individual spaces, and find surprisingly that the PageRank scores are significantly correlated to human movement both pedestrian and vehicle in four observed areas of central London. We also illustrate the fact that the topology demonstrates small world and scale free properties. The findings provide a novel justification as to why topological analysis can be used to predict human movement at a certain confidence level. We further conclude that the kind of analysis is no more than predicting a drunkard walking on a small world and scale free network.