La polysémie régulière dans WordNet

Abstract : This paper presents an analysis and modeling of polysemy in the WordNet English lexical database. It exploits the concepts hierarchy (constituted by synsets), and the gloss defining each of these concepts. The result consists of rules set which enabled us to identify in a largely automated way, with a precision close to 91%, more than 2100 synsets pairs, connected by a regular polysemy relation. Our method also allows a partial word sense disambiguation of the definition associated with these synsets.
Complete list of metadatas

Cited literature [4 references]  Display  Hide  Download

https://hal.archives-ouvertes.fr/hal-00611243
Contributor : François-Régis Chaumartin <>
Submitted on : Tuesday, July 26, 2011 - 12:12:52 AM
Last modification on : Friday, January 4, 2019 - 5:33:24 PM
Long-term archiving on : Monday, November 7, 2011 - 11:32:20 AM

File

BarqueChaumartin_polysemie_wor...
Files produced by the author(s)

Identifiers

  • HAL Id : hal-00611243, version 1

Collections

Citation

Lucie Barque, François-Régis Chaumartin. La polysémie régulière dans WordNet. TALN 2008 - 15eme conférence sur le Traitement Automatique des Langues Naturelles, Jun 2008, Avignon, France. pp. 101-108. ⟨hal-00611243⟩

Share

Metrics

Record views

733

Files downloads

331