Modèle neuronal tripartite pour la représentation de documents

Abstract : Previous work in information retrieval (IR) have shown that using evidence, such as concepts and relations, from external knowledge resources could enhance the retrieval performance. Recently, deep neural approaches have emerged as state-of-the art models for capturing word semantics that can also be efficiently injected in IR models. This paper presents a new tri-partite neural document language framework that leverages explicit knowledge to jointly constrain word, concept, and document representation learning to tackle a number of issues including polysemy and granularity mismatch. We show the effectiveness of the framework in various IR tasks including word similarity, document similarity, and document re-ranking. MOTS-CLÉS : Recherche d'information sémantique, base de connaissance, apprentissage de re-présentation.
Document type :
Conference papers
Complete list of metadatas

Cited literature [21 references]  Display  Hide  Download

https://hal.archives-ouvertes.fr/hal-01841606
Contributor : Laure Soulier <>
Submitted on : Tuesday, July 17, 2018 - 2:05:04 PM
Last modification on : Thursday, October 17, 2019 - 8:55:02 AM
Long-term archiving on : Thursday, October 18, 2018 - 2:42:27 PM

File

Coria_2018_GH.pdf
Files produced by the author(s)

Identifiers

  • HAL Id : hal-01841606, version 1

Citation

Gia-Hung Nguyen, Lynda Tamine, Laure Soulier, Nathalie Bricon-Souf. Modèle neuronal tripartite pour la représentation de documents. Conférence francophone en Recherche d'Information et Applications (CORIA 2018), May 2018, Rennes, France. ⟨hal-01841606⟩

Share

Metrics

Record views

97

Files downloads

71