Abstract
We propose a new heuristic for toponym sense disambiguation, to be used when mapping toponyms in text to ontology concepts, using techniques based on semantic similarity measures. We evaluated the proposed approach using a collection of Portuguese news articles from which the geographic entity names were extracted and then manually mapped to concepts in a geospatial ontology covering the territory of Portugal. The results suggest that using semantic similarity to disambiguate toponyms in text produces good results, in comparison with a baseline method. © 2012 Springer-Verlag.
Author supplied keywords
Cite
CITATION STYLE
Batista, D. S., Ferreira, J. D., Couto, F. M., & Silva, M. J. (2012). Toponym disambiguation using ontology-based semantic similarity. In Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics) (Vol. 7243 LNAI, pp. 179–185). https://doi.org/10.1007/978-3-642-28885-2_20
Register to see more suggestions
Mendeley helps you to discover research relevant for your work.
