Probabilistic geospatial ontologies

3rd ISWC Workshop on Uncertainty Reasoning for the Semantic Web

Sumit Sen1 ; 2

Others  :  http://CEUR-WS.org/Vol-327/paper1.pdf
PID  :  22335

来源: CEUR

【 摘 要 】

Partial knowledge about geospatial categories is critical for knowledge modelling in the geospatial domain but is beyond the scope ofconventional ontologies. Degree of overlaps between geospatial categories,especially those based on geospatial actions concepts and geospatial enitityconcepts need to be specified in ontologies. We present an approach to encodeprobabilistic information in geospatial ontologies based on the BayesOWLapproach. This paper presents a case study of using road network ontologies.Inferences within the probabilistic ontologies are discussed along withinferences across ontologies using common concepts of geospatial actionswithin each ontology. The results of machine-based mappings produced areverified with human generated mappings of concepts.