Plongement de métrique pour le calcul de similarité sémantique à l'échelle

Abstract : Plongement de métrique pour le calcul de similarité sémantique à l'échelle Résumé. In this paper, we explore the embedding of the shortest-path metrics from a knowledge base (Wordnet) into the Hamming hypercube, in order to enhance the computation performance. We show that, although an isometric embedding is untractable, it is possible to achieve good non-isometric embeddings. We report a speedup of three orders of magnitude for the task of computing Leacock and Chodorow (LCH) similarities while keeping strong correlations.
Document type :
Conference papers
Complete list of metadatas

https://hal.archives-ouvertes.fr/hal-01254852
Contributor : Julien Subercaze <>
Submitted on : Thursday, January 14, 2016 - 2:25:19 PM
Last modification on : Thursday, July 26, 2018 - 1:11:02 AM
Long-term archiving on : Saturday, April 16, 2016 - 4:20:10 AM

File

egc2016.pdf
Files produced by the author(s)

Identifiers

  • HAL Id : hal-01254852, version 1

Citation

Julien Subercaze, Christophe Gravier, Frederique Laforest. Plongement de métrique pour le calcul de similarité sémantique à l'échelle. 16 èmes Journées Francophones Extraction et Gestion des Connaissances, EGC 2016, Jan 2016, Reims, France. ⟨hal-01254852⟩

Share

Metrics

Record views

183

Files downloads

326