Trajectory Data Warehouses - Data Warehouse Systems: Design and Implementation - page 466

12.1 Mobility Data Analysis
Nowadays, with the massification of positioning devices such as GPS, we
are able to collect huge amounts of mobility data, which may be extremely
valuable in many application areas. A typical application scenario is the
analysis of the activities carried out by tourists in a city. During their
stay, tourists visit museums, parks, and several different attractions. They
also consume many services like accommodation, restaurants, shops, and so
on. From the point of view of an analyst, these tourist places and services
are denoted places of interest. A tourist trajectory consists in moving from
one place of interest to another, stopping for some time at some of them.
Data about these trajectories can be collected and analyzed, for example, to
optimize the offer of services or to plan tourist itineraries within the city.
As another example, large industrial cities with high car ownership rates are
suffering a decrease in their air quality. Normally, stations are located at
different points in these cities in order to measure air quality at regular time
intervals. It is not hard to guess that the techniques that we have studied in
this topic can be very useful for understanding and analyzing the evolution
of the quality of the air and the effects of corrective measures that the
governments may take to keep pollution below certain limits. For example, we
can analyze the trajectories followed by cars, trucks, and buses and correlate
them with the air quality measures. Or we can study the population being
exposed to heavy pollution loads and when this occurs.
In Chap. 11 , we have studied how the spatial features of objects can
be represented in databases and data warehouses. Although these spatial
features can change in time, these changes are typically considered as
discrete . For example, a parcel can be merged with another one at a certain
instant. Similarly, the borders of a state or a country can change in time.
In this chapter, we are interested in objects whose spatial features change
continuously in time. These are called moving objects . While we will deal
with moving points in this chapter, many applications must also deal with
moving regions, for example, to monitor the trajectory of polluting clouds,
or stains in sea bodies, as in our previous example. Trajectories can be
represented in a continuous or a discrete way. A continuous trajectory
is composed of the movement track of an object, occurring within a certain
interval, enriched with interpolation functions that allow us to compute,
with a reasonable degree of confidence, the spatiotemporal position of the
moving object for any instant in this interval. On the other hand, a discrete
trajectory is composed of the finite sequence of spatiotemporal positions in
a certain interval. The main difference between a discrete and a continuous
trajectory is that in the former there is no plausible interpolation function
between two points. As a typical example, consider the case of a web site