Communication Dans Un Congrès Année : 2025

Embeddings, topic models, LLM : un air de famille

Résumé

Word embeddings, topic models, LLMs: a family affair This article presents a study on terms denoting family relationships (brother, aunt, etc.) in French using three approaches: word embeddings, topic modeling, and pre-trained language models. The first two types of representations are built from the French version of Wikipedia, while the third is derived through direct interaction with ChatGPT. The aim is to compare how these three methods represent such terms, in two main ways: by evaluating them against a structural definition of family relations (in terms of features such as gender, lineage, etc.), and by comparing the topics associated with each term. These methods reveal different modes of structuring family-related vocabulary, while also underscoring the continued necessity of corpus-based and controlled analyses to obtain reliable results.

Fichier principal
Vignette du fichier
108.pdf (1.3 Mo) Télécharger le fichier
Origine Fichiers éditeurs autorisés sur une archive ouverte
Licence

Dates et versions

hal-05330625 , version 1 (26-10-2025)

Licence

Identifiants

  • HAL Id : hal-05330625 , version 1

Citer

Ludovic Tanguy, Cécile Fabre, Nabil Hathout, Lydia-Mai Ho-Dac. Embeddings, topic models, LLM : un air de famille. 20e Conférence en Recherche d’Information et Applications (CORIA) 32ème Conférence sur le Traitement Automatique des Langues Naturelles (TALN) 27ème Rencontre des Étudiants Chercheurs en Informatique pour le Traitement Automatique des Langues (RECITAL) Les 18e Rencontres Jeunes Chercheurs en RI (RJCRI), 2025, Marseille, France. pp.295-312. ⟨hal-05330625⟩
564 Consultations
103 Téléchargements

Partager

  • More