Towards Heterogeneous Resources-based Ambiguity Reduction of Sub-typed Geographic Named Entities

Mauro Gaio (1, 2) , Tien Nguyen Van (1)

1 LIUPPA - Laboratoire Informatique de l'Université de Pau et des Pays de l'Adour
2 SIGNES - Linguistic signs, grammar and meaning: computational logic for natural language

The aim of this work is to nd sub-typed Geographic Named Entities from the analysis of relations between Place Names surrounded nominal group within a specic phrasal context in a set of textual documents. The paper presents a method involving natural language processing and heterogeneous resources like gazetteers, thesauri or ontologies. The work and the results focus a French language corpus. However, the uses of quite generic lexico-syntactic patterns in pre-selected phrasal context can be tuned for others languages.

Domaines: Recherche d'information [cs.IR]

Mauro Gaio, Tien Nguyen Van. Towards Heterogeneous Resources-based Ambiguity Reduction of Sub-typed Geographic Named Entities. GeoS 2011, May 2011, Brest, France. pp.217-234. ⟨hal-00570204⟩