Global Positioning System Reference
In-Depth Information
Conclusion and Future work
In this chapter, we analyzed three different semantic similarity methods,
which are defined on the basis of information content, features'
commonality/non-commonality and combined approaches. The first
two methods are extensively used in the domain of geographic similarity
reasoning and are selected as the most representative methods in our
analysis. Accordingly, Lin , Dice , MDSM similarity methods have been
recalled and contrasted against the combined GSim method. This last
method is an ontology-centered and information content-based method that
is conceived to capture both the concept similarity within the hierarchy, and
the attribute similarity (or tuple similarity) of geographic classes. Thus, the
problem of weight assignment to concepts of reference ontology has been
considered. To this end, the frequency and probabilistic-based methods have
been analyzed. The experimental results illustrated that the GSim method
provides more reliable measures for comparing geographic classes with
respect to the selected methods.
Future research directions are the following. First, the analysis of
weight assignment to DAG (Directed Acyclic Graph)-based hierarchies
and the evaluation of results provided by the ontology-based similarity
methods. Second, the investigation of how the methods change depending
on different types of ontological relations and how it can take advantage
of these relations. In this context, it is worth investigating to what extent
the qualitative knowledge can be exploited in reasoning of concepts'
relationships. For instance, one may need to reason about how close or far
a relation is between two concepts. Similar reasoning can be made about
concepts' attributes values, like, comparing the size of a very big dig and the
size of a small lake . These studies effectively fall under the area of qualitative
spatial reasoning (Escrig and Toledo 1998), and in particular are concerned
with fuzzy spatial reasoning.
The studies on modeling fuzzy spatial relations enhance the knowledge
on spatial reasoning between elements. In (Hudelot et al. 2008), the authors
introduce an ontology of spatial relations as a support to recognizing
structures in images, which are topological, cardinal, directional and
distance relations. Topological relations between two objects are based on
notions of intersection, interior, and exterior. One of the main formalism
of such relationships, known as 9-intersections (Egenhofer 1991) uses a
partition of space into three regions for each object (its boundary, its interior
and its complement), which constitutes the basis for computing relations.
Directional relationships describe the relative position of an object with
respect to other ones, and require the space to be oriented, i.e., a reference
system. Concerning directional relations, the most used relations are related
to three axes of references: right of , left of , above , below , in front of , behind ,
Search WWH ::




Custom Search