Database Reference
In-Depth Information
10.5 Conclusions
In this chapter we have shown howmobility data mining tools can be very helpful
in supporting policy decision makers and transportation planners to answer some
very interesting research questions such as the typical access routes to a city,
the dynamics of people aggregating and scattering to/from a relevant place, the
detection of extraordinary events, and the plotting/modeling of origin destination
matrices.
In order to effectively implement and analyze policies for travel demand
management (TDM), which constitute one of the main final objectives of trans-
portation science, an increasing amount of awareness has emerged with respect
to the need for improved understanding of travel behavior. Indeed, while infor-
mation such as origin destination matrices that are derived from mobility data
mining methods may give a nice overall picture of mobility, nothing is said
about the reasons/activities behind these traffic flows. This clearly resulted in a
need for travel demand models that embody a realistic representation and under-
standing of the decision-making processes of individuals and that are responsive
to a wider range of transport policy measures. Activity-based travel analysis
approaches have received attention in recent years as a potential replacement for
trip-based approaches because they analyze travel from a theoretical perspective
that takes into account the demand for activity participation, interrelationships
among trips, and interactions among household members. In the context of the
activity-based framework, human activity is a result of actions that are moti-
vated to satisfy needs and desires of the household and its members and travel
is undertaken by individuals on their own behalf or as household members to
fulfill their needs and desires to participate in these activities. Scientific research
related to the field of activity-based modeling is motivated by the importance of
improving our understanding of human behavior on the one hand and to use this
understanding to provide better predictions of the impact of societal changes and
both travel and broader social policies on the future use of transport systems on
the other hand. Over the last decade, several of those micro-simulation models
of activity-travel demand have become operational.
Current activity-based models are based on either traditional surveys or on
full (activity) diaries to model the individual behavior of the agent in the sys-
tem. Collecting these data either in paper-and-pencil format or by means of
computer-aided technology such as small, hand-held computers is a demanding
and burdensome task for respondents. The reason for this is that data about
the principal choice dimensions underlying the simulation model have to be
collected. Typically, a temporal and spatial component always needs to be ques-
tioned. And this is exactly where larger GPS and GSM data sets, such as the ones
adopted in the research described in this chapter, could be used. However, there
is a very long way to go from raw data of individual trajectories to high-level
Search WWH ::




Custom Search