Copy the page URI to the clipboard
Hall, Mark
(2006).
URL: http://harmonisa.uni-klu.ac.at/downloads/thesis_ma...
Abstract
The Semantic Web aims to provide a more intelligent web by automatically combining information from different, heterogeneous systems. To overcome the barriers posed by this heterogeneity it is necessary to have some kind of automatic integration system. This system not only needs to provide integration services on the syntactic level, but also has to provide integration of the different semantics. This thesis introduces a model for encoding semantics that is based on cognitive principles. Building on this cognitive model a semantic similarity measure is defined that makes it possible to compare the semantics of two or more data sources in order to provide integration services. To show that the approach is usable in real-world situations it is applied to data from the land-use and land-cover domain. Evaluation of this application proves that the cognitive model and semantic similarity measure provide semantically valid results.