Geoscience Reference
In-Depth Information
Geographers have been greatly inspired by the field of statistics, especially since
the 1960s, for the “analysis” part. In particular, spatial analysis has been driven by
methodological developments based on descriptive statistics or statistical models.
The practices have since greatly diversified; first with the evolution of technologies
and computing methods, and then with the explosion of monitoring capacities and
the multiplication of data that has accompanied it. Geography has not been immune
to this development. It has found itself questioned by the whole acquisition and
processing device of all new geolocalized data. This involves the images and data
from new sensors, such as Global Positioning System (GPS): in a few years “we
have moved away from the era of the localization of information to that of the
localized information”.
1
All these developments have given rise to new needs and
new practices. These, in turn, have given birth to “schools”, and also to a mixture
between the developments carried out within different disciplines (spatial data
mining, geographical knowledge discovery, geovisualization…).
The question here is not to go further into the details of the different approaches.
If there are well-identified schools, upon which the development of environments
for specific analysis and modeling is based, the practices can then be “hybrid”. We
propose in this chapter to situate our approach in relation to this diversity of
practices. In the first step, we make explicit various challenges associated with the
chain that goes from thematic questioning to the production of knowledge. The
explanation of these challenges will highlight the different meanings that can take
the word
model
. We will stress the misunderstandings that may emerge in the
collective work, especially when geographers, archaeologists, geomaticians and
computer scientists make use of the same term without specifying it even though
they mean different implicit representations. We finally apply in a systematic
manner the various concepts introduced on rather diversified examples of empirical
issues.
2.1. From the conception of entities to their analysis of responding to thematic
issues
In this topic, we will give preference to a rather more hypothetico-deductive
approach than an inductive one. The question, formulated in the early stages, will
direct the different stages of the formalization and modeling and first of all the
choice and definition of the entities considered as relevant (choice discussed in
detail in Chapter 1). Each of the following stages corresponds to a “challenge”,
generic in the sense that it can be found fairly systematically in all the applications,
even under varied forms. These challenges fall into four categories:
1 Thierry Joliveau at the PUCA seminar on the “dynamic representations of the territories'
temporalities”, Paris, 5 February, 2014.
