Uncovering sequential patterns of travel behaviours by integrating trip-related contextual factors and personal determinants of mobility choices

The goal of this paper is to study in a unique framework some of the most relevant aspects that have been shown to influence mobility choices. On one side we consider those factors related to individual travellers, such as the socioeconomic characteristics of the household, the kind of built environment and mobility-relevant factors such as the availability of vehicles. On the other, we take into account trip-specific contextual factors such as mode choice, trip purpose and related activity patterns. By using a relatively recent and powerful development of association analysis techniques, namely sequential pattern mining, we can add a longitudinal perspective in studying the related decision making process, therefore moving from a trip-based to a tour-based approach. We apply such technique to the public use trip file of the 2009 U.S. National Household Travel Survey. Some interesting results emerged from the preliminary analyses here reported. Different ways of using motorised private means according to the gender of the traveller were highlighted, with females travelling more with others and performing more complex tours; shorter trip chains were observed for non-motorised means, whereas the patterns of use of public transport services are different according to the kind of urban environment, with longer trip tours and more differentiated users where the service offer is likely to be more attractive.