Citation

Bouyerbou H, Bechkoum K, Lepage R. Int. J. Disaster Risk Reduct. 2019; 34: 232-242.

Copyright

(Copyright © 2019, Elsevier Publishing)

DOI

10.1016/j.ijdrr.2018.11.021

PMID

unavailable

Abstract

During a catastrophic event, the International Charter11http://www.disasterscharter.org/. "Space and Major Disasters" is regularly activated and provides the rescue teams damage maps prepared by a photo-interpreter team basing on pre and post-disaster satellite images. A satellite image manual processing must be accomplished in most cases to build these maps, a complex and demanding process. Given the importance of time in such critical situations, automatic or semiautomatic tools are highly recommended. Despite the quick treatment presented by automatic processing, it usually presents a semantic gap issue. Our aim is to express expert knowledge using a well-defined knowledge representation method: ontologies and make semantics explicit in geographic and remote sensing applications by taking the ontology advantages in knowledge representation, expression, and knowledge discovery. This research focuses on the design and implementation of a comprehensive geographic ontology in the case of major disasters, that we named GEO-MD, and illustrates its application in the case of Haiti 2010 earthquake.

RESULTS show how the ontology integration reduces the semantic gap and improves the automatic classification accuracy.

Language: en

Keywords

Information retrieval; Major disasters; Ontology; Ontology web language (OWL); Reasoning; Semantics