A Method to Model Population-Wide Social Networks for Large Sc1 aleActivity-Travel Micro-Simulation

摘要

Social-leisure activities account for an important and increasing segment of travel in modern societies. Yet, these activities are least understood in current activity-based models of travel demand. In this paper we propose a model to generate population-wide social networks that in the context of large-scale micro-simulation of travel demand provide a basis for modeling social interactions. The proposed model consists of a friendship formation model formulated in the RUM framework, and a component to simulate the network in a population. We show how the friendship model can be estimated by loglikelihood methods on observations of personal networks. In an application to the Swiss context, we demonstrate the estimation and ability of the model to reproduce relevant characteristics of networks, including for the first time simultaneously geographic distance, attribute similarity (homophily), size of personal networks (degree distribution) and clustering (transitivity). We conclude that the model, in combination with current methods to generate synthetic populations, offers a basis to model social-leisure activities and associated travel in more rigorous behavioral ways than previously possible.