STUDYING THE PATTERNS OF USE OF TRANSPORT MODES THROUGH DATA MINING: AN APPLICATION TO THE U.S. NATIONAL HOUSEHOLD TRAVEL SURVEY DATASET

摘要

Travel-related data collection activities require high amounts of financial and human resources tobe successfully carried out. In a context where the available resources are scarce, there is a needto exploit the information that is hidden in these dataset, to increase their added value and gainsupport among decision makers not to discontinue such efforts. The present research assesses theuse of a data mining technique, Association Analysis, to better understand the patterns of modeuses from the 2009 U.S. National Household Travel Survey. Only variables related to self-reported levels of use of the different transportation means are considered, along with thoseuseful to the socioeconomic characterization of the respondents. It has been possible to mineassociation rules that potentially show in economic terms a substitution effect between cars andpublic transportation, whereas such effect was not observed between public transportation andnon-motorized modes (bike, feet). This is a policy relevant finding, since transit marketing shouldbe targeted to car drivers rather than to bikers or walkers to really improve the environmentalperformances of any transportation system. Modal diversion from car to transit is seldomobserved in practice, given the competitive advantage of private modes that has been extensivelydiscussed in the literature. However, if we control for such factor, then our results suggest thatmodal diversion should mainly occur from cars to transit, rather than from non-motorized modesto transit.