The International Archives of the Photogrammetry. Remote Sensing and Spatial Information Sciences. Vol. XXXVII. Part B4. Beijing 2008
1096
identifier is not task of the ontology because of the inherent
given deictic references.
Distance aspects also have to be considered for characterizing
spatial scenes and analyzing spatial descriptions. As a basic
principle quantitative and qualitative descriptions of distance
parameters have to be distinguished. Quantitative descriptions
are based on units of lengths (meters) as well as units in time
(minutes). According to that a quantitative description of a
distance is always an isotropy. In contrast qualitative distance
descriptions are terms like quite near or far away. Such
descriptions are anisotropy because they depend on the
observer, position, size, visibility, dominance and a lot of other
features. Thus a quantitative representation of a qualitative
distance can be achieved by classification with fuzzy
membership functions (cf. Guesgen, 2002).
5. CONCLUSION
Information retrieval in the disaster management domain is
based on verbally given descriptions of spatial scenes. Spatial
scenes here are quite dynamic situations and require that
topology, neighborhood, orientation as well as distance aspects
are considered. For a semantically correct representation of the
required general and context knowledge, complex structures are
necessary, as given by ontologies.
Such ontology provides a spatial reasoning process and makes
different types of reasoning possible. By modeling relations
between classes, topological relations are directly inferable. To
amplify this type of reasoning some additional relations have to
be identified by a spatial reasoning algorithm. Therefore the
intersection model (IM) and the region connection calculus
(RCC) are evaluated. Both algorithms make reasoning with
respect to the needed set of fl R topological relations possible.
But using the IM algorithm seems practicable because this
algorithm is already implemented in most GIS components.
For the evaluation of neighborhood relations also different
methods are analyzed. The distance based definition is
compared to a triangulation method, whereas the best method
for representation and identification with respect to the systems
requirement is the Delaunay triangulation. Based on object
center points, a natural neighborhood in terms of graph theory is
given by an edge of the connecting triangle.
ACKNOWLEDGEMENTS
The work of Christian Lucas has been funded by the Deutsche
Forschungsgemeinschaft (DFG), project no. BA 686/16
“Abstraction of Graphically and Verbally Represented
Geoinformation”.
Cicerone, S. and Clementini, E.: Efficient Estimation of
Qualitative Topological Relations based on the Weighted
Walkthroughs Model, Geolnformatica, Volume 7, pp. 211-
227,2003.
Egenhofer, M.-J. and Herring, J.-R.: Categorizing binary
topological relations between regions, lines, and points in
geographic databases. Technical report, Department of
Surveying Engineering, University of Maine, 1991.
Frank, A.U.: Formal models for cognition - taxonomy of spatial
location description and frames of reference. Spatial
Cognition: An Interdisciplinary Approach to Representing
and Processing Spatial Knowledge, pp. 293-312, 1998.
Gold, C.M.: The Meaning of ’’Neighbour”. Theories and
Methods of Spatio-Temporal Reasoning in Geographic
Space, Volume 639, pp. 220-235, 1992.
Guesgen, H.-W.: Reasoning About Distance Based on Fuzzy
Sets. Applied Intelligence, Volume 17/3, Kluwer Academic
Publ., Boston, pp. 265-270, 2002.
Grigni, M., Papadias, D. and Papadimitriou. C,: Topological
inference. In IJCAI-95, pp. 901-906, 1995.
Hlavaty, T. and Skala, V.: Combinatorics and Triangulations.
Computational Science and Its Applications, Lecture Notes
in Computer Science, Volume 3045, Springer, pp. 81-89,
2002.
Hernández, D.: Qualitative Representation of Spatial
Knowledge. Volume, Lecture Notes in Artificial
Intelligence, Volume 804, Springer, 1994.
Koch, A.: Semantische Integration von zweidimensionalen GIS-
Daten und Digitalen Geländemodellen. Dissertation, Verlag
der Bayerischen Akademie der Wissenschaften, 2007.
Lucas, C., Werder, S. and Bähr, H.-P.: Information Mining for
Disaster Management. International Archives of
Photogrammetry, Remote Sensing and Spatial Information
Science, Volume 36, Part 3, pp. 75-80, 2007.
Renz, J.: Qualitative Spatial Reasoning with Topological
Information. Lecture Notes in Artificial Intelligence,
Volume 2293, Springer, 2002.
REFERENCES
Barkowsky, T.: Mental representation and processing of
geographic knowledge. Lecture Notes in Artificial
Intelligence, Volume 2541, Springer, 2002.
Billen, R. and Clementini, E.: A Model for Ternary Projective
Relations between Regions. Advances in Database
Technology, Lecture Notes in Computer Science, Volume
2993, Springer, 2004.