Hanna, S; (2019) Random walks on weighted graphs: centrality measures or simple agents. In: The 12th International Space Syntax Symposium. Space Syntax Symposium: Beijing, China.

Preview

Text 431-2_Hanna_RandomWalks.pdf - Published Version Download (2MB) | Preview

Abstract

Two new methods of graph segment analysis are introduced and tested, which use a random walk as a theoretical and practical basis. The first, random walk betweenness (RWB), is a measure of betweenness centrality, similar to choice, and the second is a simple calculation of a random walk on a graph with similarities to visual agent simulation. To implement these, two methodological innovations are made: angular weighting for probability of turns and a dual node representation of segments. The two methods are tested for their correlation with both pedestrian and vehicular movement and compared with segment angular choice. RWB is seen not to be as effective a prediction of human movement, but simple random walks appear approximately as effective as choice, and faster to compute. Practically, this yields a substantially faster algorithm for movement prediction than previously exists. Theoretically, it suggests a much simpler model of agency: that navigation may be more opportunistic than optimised. To the extent the latter is valid it may prove valuable to a more complete theory of movement and a bridge between Space Syntax and related fields of mathematical graph theory.

Download activity - last month

Export as JSONXMLCSV

Download activity - last 12 months

Export as XMLJSONCSV

Downloads by country - last 12 months

Export as CSVJSONXML

Archive Staff Only

View Item