Multi-criteria modelling and clustering of spatial information
Abstract:To improve the formal integration of verbally given spatial information in a geographical information system, a methodology for converting such information to a geographical representation was developed. This was applied to examples of a spatial-related analysis of natural language text messages in the domain of disaster management. This article presents an approach for such a conversion by using a suitable knowledge representation as well as formal modelling structures. The structures are provided by an ontology-supported knowledge base. With respect to spatial uncertainty, a formal representation based on possibility theory was defined. Moreover, a cluster algorithm was developed for handling information of different sources and building a spatial context.
Document Type: Research Article
Affiliations: Institute of Photogrammetry and Remote Sensing, Karlsruhe Institute of Technology (KIT), Karlsruhe, Germany
Publication date: 2012-10-01