Geoscience Reference
In-Depth Information
(or properties), and relationships (or relations among class members).” Knowledge
bases are obtained by feeding ontologies with knowledge about particular ele-
ments (usually called individuals), their attributes and their relationships to other
individuals. Ontologies (and knowledge bases) differ from databases in the sense
that “the languages of ontologies are closer in expressive power to first-order
logic than languages used to model databases. For this reason, ontologies are said
to be at the “semantic” level, whereas database schema are models of data at the
“logical” or “physical” level” (Gruber
2009
). Moreover such a formalized ontology
can be associated to a reasoner that can perform some computational reasoning.
In our case, an ontology-based approach enables (1) a formal representation of
existing (but scattered) knowledge about 3D visualization techniques and (2) logi-
cal reasoning that allows some computation.
2.2 Visualization Ontologies or Taxonomies
Using 3D city models for helping decision making in the urban domain is related
to tasks performed by the user of these models and for which he/she has to navi-
gate in or over the 3D model. Typical tasks include evaluation of urban projects
in terms of quality of life (including visual aspects), evaluation of the impact of
projects on the urban landscape and on other factors. Such tasks imply the visuali-
zation of data that can originate from different fields like transport, construction,
air quality, noise, etc.; be of different kinds such as quantitative measures of noise
or qualitative soundscapes; take different forms, from structured data provided
by geographical information systems to unstructured documents; have different
scales (city as a whole, buildings, etc.); be not directly georeferenced (legal text
for example) although they have a spatiotemporal coverage. What has to be visu-
alized is not only data (such as real or integer numbers) but information (such as
pollutant values measured at a given height above a street, counting of pedestrians
having passed upon given places, etc.).
Different classifications, terminologies, taxonomies or ontologies have been
defined in the field of visualization, with different aims. Gao et al. (
2008
) pre-
sented the design of a visualization ontology, which aims at providing more
semantics for the discovery of visualization services. The Top Level Visualization
Ontology (TLVO) defined by Brodlie and Noor (
2007
) aims at providing a com-
mon vocabulary to describe visualization data, processes and products. Based on
an analysis of visualization taxonomies and on more recent work in visualization
ontologies, Morell Pérez et al. (
2011
) propose some modifications to the TLVO
in order to better represent the visualization process and data models. Voigt and
Polowinski (
2011
) aim at developing a unifying ontology applicable in visualiza-
tion systems. Bazargan and Falquet (
2009
) have proposed to use description logics
to represent the usability of techniques in a given context and to reason about it.
More recently, Voigt et al. (
2012
) have created a visualization ontology that
supports a recommendation system for the selection of visualization components.
