DBnary2Vec: Preliminary Study on Lexical Embeddings for Downstream NLP Tasks - Archive ouverte HAL Accéder directement au contenu
Communication Dans Un Congrès Année : 2023

DBnary2Vec: Preliminary Study on Lexical Embeddings for Downstream NLP Tasks

Résumé

In this preliminary study, we experiment with the use of DBnary, a big lexical knowledge graph, to create word embeddings that could be used in NLP downstream tasks. Our gamble is that word embeddings created from lexical data (instead of language corpora) may exhibit less biases while still being usable as the first layer of deep learning approaches to NLP tasks. We tried very basic method of embedding creation from lexical graph and evaluate (1) the intrinsic performance of the created embeddings on word similarity and word analogy test sets and their extrinsic quality in POS tagging and NER downstream tasks, along with (2) the biases they may exhibit. Such embeddings show promising performances outperforming word2vec on few specific tasks, while still not on par on most others, but we confirm that they exhibit less bias overall.
Fichier principal
Vignette du fichier
DL4LD_2023_DBnary (1).pdf (305.78 Ko) Télécharger le fichier
Origine : Fichiers produits par l'(les) auteur(s)
Licence : CC BY - Paternité

Dates et versions

hal-04192640 , version 1 (31-08-2023)

Licence

Paternité

Identifiants

  • HAL Id : hal-04192640 , version 1

Citer

Nakanyseth Vuth, Gilles Serasset. DBnary2Vec: Preliminary Study on Lexical Embeddings for Downstream NLP Tasks. 3rd Workshop DL4LD: Addressing Deep Learning, Relation Extraction, and Linguistic Data with a Case Study on The Bigger Analogy Test Set (BATS), Sep 2023, Vienna, Austria. ⟨hal-04192640⟩
21 Consultations
20 Téléchargements

Partager

Gmail Facebook X LinkedIn More