Geoscience Reference
In-Depth Information
As OWL can handle only textual concepts, it needs to be revised in order
to describe visual aspects as well. (Karam et al. 2010) had well explained
our scenario. A more invasive suggestion is to propose an extension of
OWL standard; we named it CartOWL in order to describe in a dedicated
and organized file all the visual concepts and their relationships of equiva-
lence and inclusion.
Once the local ontologies corresponding to the LBS providers' carto-
graphic visual concepts are generated, then the matching/alignment step
should start.
The full prototype will be able to parse the CartOWL output files and align
them towards one reference knowledge base (domain ontology) so that we
can ensure map conflation results on mobile devices.
Belief function must be applied as well through CartOWL in order to
achieve the best compromise between the domain ontology and other con-
straints that may interfere such as the user's profile (nationality, map pref-
erence, age, etc.), the context of his geographic zone, the graphical semiology
rules and color contrasts v/s visibility, the device limitations and the need
for generalization, adaptation and dynamic maps, etc. So, in order to
prioritize visual attributes from one provider among others, highest
weights will be assigned to them as per the belief theory in the CartOWL
tags. Psycho-cognitive test for efficient icons-recognition, without legends,
will help us to assign such degrees of preference or weight and prioritize
an icon among many representing the same service.
Some screenshots of our building/matching prototype are shown below
(
Fig. 4
):
