Nowadays, there is a growing interest within the geographic information system community to find the location of geographic data through the Internet and to know the features and the possibilities of these data. In order to make this information easily accessible both to trained users and to the general public it is necessary to implement a specific infrastructure of geographical information providing advanced classification and searching services. The ISIGIS project is aimed at creating such an infrastructure in Spain. In this context, when a geographic information Catalog collects the description of geospatial information from different contexts, it is not reasonable to assume that a particular user outside the community where the data were created will describe the sought information by means of the same terms or keywords in the Catalog. Then, the search service of the Catalog cannot be a mere keyword recognition process. It should be able to understand the sense of the user's vocabulary and to link these meanings to the terms in the thesaurus (or thesauri) used in the catalog. In this paper we present an heuristic method based on a voting system to tackle this disambiguation problem in geographic Catalogs related to the ISIGIS project. Our proposal can be considered as an unsupervised disambiguation method based on the hierarchical structure of both WordNet and the thesauri.
[1]
Eneko Agirre,et al.
Word Sense Disambiguation using Conceptual Density
,
1996,
COLING.
[2]
George A. Miller,et al.
Introduction to WordNet: An On-line Lexical Database
,
1990
.
[3]
Philip Resnik,et al.
A Perspective on Word Sense Disambiguation Methods and Their Evaluation
,
2002
.
[4]
Philip Resnik,et al.
Disambiguating Noun Groupings with Respect to Wordnet Senses
,
1995,
VLC@ACL.
[5]
Kenneth D. Forbus,et al.
Using a Geographic Information System for Qualitative Spatial Reasoning about Trafficability
,
1999
.
[6]
Michael Gould,et al.
Aspectos tecnológicos de la creación de una Infraestructura Nacional Española de Información Geográfica
,
2001
.
[7]
David Yarowsky,et al.
A method for disambiguating word senses in a large corpus
,
1992,
Comput. Humanit..
[8]
Carol Peters,et al.
Applying EuroWordNet to Cross-Language Text Retrieval
,
1998,
Comput. Humanit..