A user-centric model of connectivity in street networks

Modelling street network connectivity is a fundamental research problem in transportation science. Here, we argue that the connectivity of a street network cannot be defined independently from the users of that street network. That is, different users will experience different levels of connectivity depending on their corresponding travel behaviour. In this work, we propose a model of connectivity that is defined with respect to a user’s travel behaviour within a given street network. We demonstrate that many real-world problems can be posed as instances of an optimisation problem defined with respect to this model. This includes optimising a user’s home location with respect to their connectivity and optimising the location of a facility with respect to the connectivity of its users. We prove the above optimisation problem is NP-hard and present an integer programming solution that scales to large problem instances.