Approche supervisée à base de cellules LSTM bidirectionnelles pour la désambiguïsation lexicale

Abstract : In word sense disambiguation, there are still few usages of neural networks. This direction is very promiseful however, the results obtained by these first systems being systematically in the top of the evaluation campaigns, with an improvement gap which seems still high. We present in this paper a new architecture based on neural networks for word sense disambiguation. Our system is at the same time less difficult to train than existing neural networks, and it obtains state of the art results on most evaluation tasks in English. The focus is on the reproducibility of our systems and our results, through the use of a word embeddings model, training corpora and evaluation corpora freely accessible.
Liste complète des métadonnées

Cited literature [29 references]  Display  Hide  Download

https://hal.archives-ouvertes.fr/hal-01781183
Contributor : Didier Schwab <>
Submitted on : Sunday, April 29, 2018 - 5:12:26 PM
Last modification on : Tuesday, February 12, 2019 - 1:31:31 AM
Document(s) archivé(s) le : Thursday, September 20, 2018 - 1:38:00 AM

File

TALN2018_Vial-et-al.pdf
Files produced by the author(s)

Identifiers

  • HAL Id : hal-01781183, version 1

Collections

Relations

Citation

Loïc Vial, Benjamin Lecouteux, Didier Schwab. Approche supervisée à base de cellules LSTM bidirectionnelles pour la désambiguïsation lexicale. 25e conférence sur le Traitement Automatique des Langues Naturelles, May 2018, Rennes, France. ⟨hal-01781183⟩

Share

Metrics

Record views

142

Files downloads

150