Database Reference
In-Depth Information
process to cope with large-scale data sets. Trajectory data are not an exception
to this phenomenon. The popularization of tracking technologies (e.g., GPS-
enabled mobile devices) had as a consequence that huge amounts of trajectory
data are being continuously collected. Therefore, all the analysis methods and
software tools that have been presented in this topic must be redesigned in order
to scale up for much larger data sets, as most of the methods are quite restrictive
with respect to data volume.
A related problem in this respect pertains to developing methods for stream-
ing trajectory data. In many real-world applications (e.g., telecommunications,
clickstream monitoring, sensor networks, traffic monitoring), data take the form
of continuous data streams rather than finite stored data sets. These applications
require long-running, continuous queries and analyses as opposed to single-time
ones. Many aspects of data management and processing need to be reconsidered
in this new setting and stream databases were developed as a possible solution
for this. In a similar way to large-scale processing, new methods and tools have
to be designed to enable stream-based processing of trajectory data.
Mobility Engineering
Definitely, more systematic exploration and experimentation is necessary to con-
solidate the theories and tools that this topic has presented. Most of the experi-
mentations that allowed researchers to assess their results have been carried out
on an ad hoc basis and were limited to the data set available to the researchers.
Systematic exploration of the applicability of an approach to mobility manage-
ment remains to be done. Validating an approach for large-scale usage in the
real application world needs repeated testing with varying parameters, varying
techniques, and varying data sets. Among others, ground truth benchmarks need
to be developed to create better possibilities to assess the value and portabil-
ity of algorithms. Moreover, all involved tools and facilities will have to reach
online availability to enable continuous analysis of and feedback for ongoing
trajectories.
Turning research into engineering represents a huge challenge and calls for
a strong cooperation among research, industry, users, and public authorities.
The ultimate goal is to be able to develop general-purpose packages that will
allow, for example, to translate GPS tracks into semantic trajectories. Similarly,
general platforms should enable the tuning of the parameters for all tools that
create, manage, and manipulate a trajectory data set. It is a long way to go, but
as shown in this topic we hopefully are on the right track.
