Database Reference
In-Depth Information
already mentioned in Chapter 15), which could potentially offer insights in how
people move from one state to another over time. In the end, however, it is the
rapid development of positioning technologies such as GPS, and the growing
penetration of these technologies in mobile devices, such as car kits and mobile
phones, which acts as the main catalyst in a new and burgeoning research area.
After all, these devices can be regarded as very good proxies for capturing human
mobility.
As a result, there has been a rapid increase in the amount of mobility data
sets. As these data sets tend to be large, the sheer volume of data confronts
researchers with difficulties in extracting interesting and relevant knowledge.
While the importance of this issue - often used in paradigms such as “the data
avalanche” - is undeniable, it should also be stressed that human mobility is
not always as easily measurable as might be conceived at first sight. First and
foremost, persons can move around in a variety of ways. As more and more
vehicles are equipped with GPS navigation kits, the movement data from these
vehicles are already used for purposes such as the real-time monitoring and
prediction of traffic jams. Capturing the movements of cyclists and pedestrians,
however, is already significantly more difficult. Mobile phones usually remain
very close to their owners at all times, so they are the most obvious candidate
proxies. Because mobile operators keep records of telephone calls making use
of their cell towers, it is possible to reconstruct movements of phones by mining
their call logs. This methodology - usually called “mobile positioning” - delivers
very large mobility data sets that have already been used to study regional
movements. GPS loggers carried around by a test audience form an alternative
way of measuring themovements of people. Because the resulting trajectories are
usually very accurate and participants can also be surveyed before or after their
cooperation, this method is becoming increasingly popular among scientists.
Both methodologies, however, have their deficiencies. First, the cooperation
with mobile operators for mobile positioning data sets has proven to be diffi-
cult. More importantly, the spatial accuracy of this methodology (at best a few
hundred meters in urban settings) is insufficient for studying human mobility on
smaller scales. Alternatively, the distribution and recollection of GPS loggers
among a test audience is labor intensive and possibly expensive, which will auto-
matically result in a smaller sample size. Additionally, research projects making
use of this technology will essentially be limited to outdoor environments where
shadowing due to dense urban environments can potentially lower data quality as
well.
The difficult nature of capturing human mobility on smaller (in this con-
text subregional) scales shows that, despite the undeniable data avalanche
confronting researchers, there remain challenges in capturing movement data
besides processing them. In short, there is a need for a methodology that can
measure human movement on a small scale in a cost- and labor-effective way, in
